AN ABSTRACT OF THE DISSERTATION OF
Candice Michelle Weems for the degree of Doctor of Philosophy in Geography
presented on March 18, 2016.
Title: Examining the Spatial Distribution of Park Access and Trajectories of
Gentrification in Seattle, Washington 1990 - 2010.
Abstract approved:
______________________________________________________
Julia A. Jones
This dissertation examines the spatial distribution of park access by type in
relation to trajectories of gentrification in Seattle from 1990 to 2010. The dissertation
includes 5 Chapters. The first chapter provides an overview of the literature that
motivated this research. The second, third and fourth chapters are research papers
that seek to address the overall question: What is the relationship between park
access and trajectories of gentrification? Chapter 5 is an overall conclusion. This
study tests how the urban parks, shape the communities around them and focuses on
three main objectives: (1) To gain a better understanding of socioeconomic and racial
factors that are associated with gentrification in Seattle, Washington, (2) To develop a
transferrable approach for assessing trajectories of gentrification in urban areas, and
(3) To determine how the spatial distribution of different park types and community
investment in parks are associated with gentrification. The research is motivated by
the environmental justice literature and seeks to explore issues as they affect where
people live and how they interact with parks in a large urban area. Little is known
about how environmental amenities such as parks interact with socio-economic and
racial characteristics of neighborhoods over time. Chapter 2 , examined the
relationships among park system development over time, park typologies, and socio-
economic and racial patterns, in order to understand how park system evolution is
related to environmental justice. The results suggest that park access differed among
socio-economic groups, and by park type. Significantly greater proportions of
African Americans and Asians were found in census block groups with below-median
levels of education, income, and home value, which in turn had access to significantly
larger numbers and more area of recreation parks. In contrast, census block groups
with above-median income, home value and education, which had significantly fewer
minorities, had access to greater numbers and more area of natural passive and
multiuse natural parks. The patterns did not change much over the study period
suggesting that park development had the effect of maintaining differences in park
access. This work raises the question of whether investments in park improvements
led or followed the changes in educational attainment, home value, and income,
suggesting that further work is needed to test hypotheses of environmental
gentrification in Seattle. Chapter 3, used available digital spatial data including
census data as well as spatially explicit mixed media coverage to map gentrification
in order to test hypotheses about causes and consequences of the process.
Gentrification was defined as change from below to above the median value for
education, income, and home value in a census block group over the period 1990 to
2000 or 2000 to 2010. These patterns were compared to changes in proportions of
minorities and perceptions of gentrification in the media over those time periods.
Gentrification was clustered in central, north and west Seattle. In Central Seattle
these changes were accompanied by large declines in the proportion of African
Americans in the population. However, gentrification in terms of education, income,
and home value also occurred in areas that did not have large proportions of
minorities. Media perceptions of gentrification were reported in places that met the
definition of gentrification, but also in areas experiencing changes in minorities, but
which did not yet meet the definition of change in terms of education, home value,
and income. The findings of this study suggest that gentrification is multifaceted,
multidimensional and spatially contagious process that does not always lead to
displacement of minorities. More work is needed to understand the causes of
gentrification especially with regard to the role of urban parks. Chapter 4 examined
the relationships between spatial patterns of change and investment in urban parks
and processes of gentrification overtime. A spatial analysis was conducted using GIS
to test relationships between changes in socioeconomic characteristics and changes in
parks from 1990 to 2010. Changes in access to three different park types (recreation,
natural passive, and multiuse natural) were compared to changes in socioeconomic
characteristics from 1990 to 2000 and 2000 to 2010. Spatial data on communityinitiated investment in parks were also compared to spatial patterns of change in park
access and socioeconomic characteristics. The results suggest that gentrification was
associated with locations where the number and acreage of recreational and natural
passive parks increased from 1990 to 2000, and where high levels of park investment
had occurred from 2000 to 2010. Therefore, changes in park access and investment
influence the character of neighborhoods over time. The results also showed that
moderate and low values of park investment were less associated with gentrification.
The findings ultimately contribute to the understanding of resilience of a community
and the capacity to absorb changes without being pushed into a gentrified regime.
Lastly, Chapter 5 provides an overall conclusion to this dissertation and provides
suggestions for future research.
Keywords: GIS, park access, gentrification, environmental gentrification, community
investment
©Copyright by Candice Michelle Weems
March 18, 2016
All Rights Reserved
Examining the Spatial distribution of Park Access and Trajectories of Gentrification
in Seattle, Washington 1990 - 2010
by
Candice Michelle Weems
A DISSERTATION
submitted to
Oregon State University
in partial fulfillment of
the requirements for the
degree of
Doctor of Philosophy
Presented March 18, 2016
Commencement June 2016
Doctor of Philosophy dissertation of Candice Michelle Weems presented on March
18, 2016
APPROVED:
Major Professor, representing Geography
Dean of the College of Earth, Ocean, and Atmospheric Sciences
Dean of the Graduate School
I understand that my dissertation will become part of the permanent collection of
Oregon State University libraries. My signature below authorizes release of my
dissertation to any reader upon request.
Candice Michelle Weems, Author
ACKNOWLEDGEMENTS
This has been a wonderful journey. I am thankful to God who has brought
this far and will continue to take me places. Thank you to all that have contributed
their time and effort to help me become the scholar that I am.
I would like to take this time thank key players in this dissertation and my life.
Firstly, I would like to thank my advisor, Dr. Julia A. Jones. She has never doubted
my abilities. She took me under her wing back in 2010 and has taught me so much,
cared for me and my family and has gone above and beyond the call of an advisor.
She is top notch and I am ever grateful for her kindness, generosity and guidance.
Next I would like to thank my committee members, Dr. Mary Santelmann, Dr. Lori
Cramer, Dr. Jenna Tilt, and Dr. Christine Olsen (GCR). My committee members
have encouraged me and taught me so much over the years. I appreciate their
willingness to sit with me and guide and challenge me through my dissertation and
provide me with thoughtful insight.
My family has played a key role in my journey throughout life and graduate
school. To my mother, Hannah Weems, father, Melvin Weems, brothers (Mark and
Michael) and sister (Melanie), THANK YOU! You have encouraged me from day
one. My family has been the key to my success and I am grateful for their support
and never failing love. To my husband, Daniel Landry, thank you for taking this
journey with me. I appreciate your support and continued love and encouragement. I
am also thankful for my daughter, Hannah who has filled my heart with joy and
laughter. She is the apple of my eye.
Lastly, I would like to thank my extended family: my friends (Jen Vielleux,
Alison Doniger, Julie Watson, Julie Klath, Kim Ogren, Lucy Romeo, Jacob Petersen
Perlman, Laura Hardin, Melva Trevino-Peña, Katherine Jones, Melissa McCracken,
Dori Dick - and a host of other friends). I love and care for you all. Together through
the years we have done it all! We will forever remain friends through the thick and
thin. And to my best friend, Shiloh Sundstrom, I love you and will forever miss you.
TABLE OF CONTENTS
Page
Chapter 1 Introduction……………..………………………………………………… 1
1. Conceptual Framework………………………………………………..…... 2
2. Environmental Justice...…………………………………………………… 4
3. Trajectories of Gentrification……………………………………………… 9
3.1 Cultural theories of Gentrification...………………………………….... 9
3.2 Economic theories of Gentrification………………………………….. 10
3.3 Environmental Gentrification..........………………………………….. 11
4. Park Planning in Seattle……………………………………………..…… 12
5. Park Typology…………………………………………………………… 14
6. Conclusion……………………………………………………………….. 14
7. References………….……………………………………………..……… 15
Chapter 2 Access to parks in Seattle: relationship of park type to socioeconomic
variables, 1990 – 2010……………………………………………………………… 20
1 Introduction and Research Questions……………………………………...22
1.1 Park typologies…………………………………………………………22
1.2 Evolution of park systems over time – the role of the government……24
1.3 Park access and environmental justice………………..…………….….24
1.4. Research questions: Park system evolution, park types, and
environmental justice…..…….…………………………………………….25
1.5 Research hypotheses………………………………………………...…26
2 Study Site…..………………………………………………………………28
3. Data Sources and Methods……………………………………………...…31
TABLE OF CONTENTS (Continued)
Page
3.1 Socioeconomic data sources…………………………………………...31
3.2 Park data sources.………………………………………………………33
3.3 Defining above-and-below median census block groups…..………….34
3.4 Park types, distribution and access………………………………...…..34
3.5 Statistical Analysis……………………………………...…..………….36
4 Results ………………………………………………………………………36
4.1 Park distribution by type and changes in spatial distribution of parks
over time…………………………………………………………………..36
4.2 Locations of above and below median census block groups…..……...38
4.3 Relationships among park area, park type, home value income, and
education………………………………………………………………….43
4.4 Access to Recreation Parks and change over time by education, home
value and income...................................................................................…...44
4.5 Access to Natural Passive Parks and change over time by education,
home value and income…….........................…………………..………....45
4.6 Access to Multiuse Parks and change overtime by education, home value
and income...................................………………………………………....47
4.7 Trends in park access over time by socioeconomic
group……………….........................………………………………………49
4.8 Relationship of park access to race…………………...……..…………50
5 Discussion………………………………………………………………….52
5.1 Park access and socioeconomic variables in Seattle…………………..52
5.2 Methodological issues………...………….…………………………….52
5.3 Changes in park distribution over time – the role of park
investment…………………………………………………………............53
TABLE OF CONTENTS (Continued)
Page
5.4 Disparities in park access and environmental justice………………….55
5.5 Changes in park access and implications for gentrification……………56
5.6 Implications for park planning…………………………………………56
6. Conclusion…………………………………………….………….……….57
7. References…………………………………………….…………………..59
8. Appendix………………………………………………………………….62
Chapter 3 Spatial patterns of gentrification in Seattle: multifaceted, dynamic
interactions…………..……………………………………………………………….65
1 Introduction and Research Questions……………………………………...66
2 Study Site…..………………………………………………………………68
3. Data Sources and Methods……………………………………………...…68
3.1 Data sources…………………………….……………………………...68
3.2 Spatial data analysis……………………………………………………70
3.3 Mapping gentrified census block groups……………….…..………….71
3.4 Change in African American populations………..………………...…..73
3.5 Content analysis on select neighborhoods of gentrified census block
groups……….……………………………………………....…..………….73
4 Results ……………………………………………………………………74
4.1 Changes in spatial patterns of education, home value and household
income….………………………………………………………………..74
TABLE OF CONTENTS (Continued)
Page
4.2 Spatial patterns of socioeconomic change and gentrification, 1990 to
2000……………….…………………………………………….…..……...81
4.3 Spatial patterns of socioeconomic change and gentrification, 2000 to
2010………………………………………………………………….…….84
4.4 Racial changes in gentrified census block groups, 1990 –
2000……………....................................................................................…...86
4.5 Racial changes in gentrified census block groups, 2000 –
2010………………………….........................…………………..………....89
4.6 Relationships of racial change to
gentrification.................................………………………………………....95
4.7 Within neighborhood spatial and temporal patterns of gentrification and
racial change.……………................………………………………………98
4.8 Content analysis on select neighborhoods of gentrified census block
groups 1990 – 2010.……………………………………………..…….….100
4.8.1 Demographic shifts……………………………..………………...101
4.8.2 Sense of Community (or lack thereof)……………………………102
4.8.3 Economic and development booms…………………………..…..103
5 Discussion………………………………………………………………...103
5.1 Hypothesis 1………………………………………………………….104
5.2 Hypothesis 2………………………………………………………….105
5.3 Hypothesis 3………………………………………………………….106
5.4 Hypothesis 4………………………………………………………….107
5.5 Methodological issues…………………………………………...……107
6. Conclusion………………………………………………………..……...108
TABLE OF CONTENTS (Continued)
Page
7. Literature Cited………………………………………………………….109
Chapter 4 Gentrification and changes in park access and investment in Seattle, 19902010……..…………..……………………………………………………..……….113
1 Introduction and Research Questions………………………………...…..114
2 Study Site…..…………………………………………………………….116
3. Data Sources and Methods……………………………………………... 117
3.1 Socioeconomic data sources…………....………………………….....117
3.2 Data on park investments……………………………………………..118
3.3 Determine changes in park access…………………….…..………….120
3.4 Identify gentrified census block groups………..………………...…..120
3.5 Define environmental gentrification and how environmental
gentrification will be tested……….…..…………………....…..…………120
4 Results ……………………………………………………………………120
4.1 Park investment vs.
gentrification……………………………………………………………...120
4.2 Changes in park access and
gentrification……….……………………………………………..…........124
4.2.1 Changes in park access and gentrification –recreation parks….....124
4.2.2 Changes in park access and gentrification – natural passive
parks……………………………………………………………………127
4.2.3 Changes in park access and gentrification – multiuse parks…......129
4.3 Park investment, changes in park access and
gentrification……….……………………………………………..…........132
TABLE OF CONTENTS (Continued)
Page
4.3.1 Park investment, changes in park access and gentrification –
recreation parks…………………….………………………………......132
4.3.2 Park investment, changes in park access and gentrification – natural
passive parks…………………….………………………………..........135
4.3.3 Park investment, changes in park access and gentrification –
multiuse parks…......................................................................................137
5. Discussion………………………………………………………..……....140
6. Conclusion………………………………………………………..……...142
7. Literature Cited………………………………………………………….143
Chapter 5 Conclusion…………….………………………………………..……….145
LIST OF FIGURES
Figure
Page
1.1 The society and land use model………………………..…………………...….. 3
2.1 Census block groups in the city of
Seattle………............................................................……………………………….. 30
2.2 Directional Distribution of parky types added to park system 1990-201…..…... 38
2.3 The spatial distribution of census block groups with household income above or
below the median household income in Seattle, 1990, 2000, and 2010………..........40
2.4 The spatial distribution of census block groups with home value above or below
the median home value in Seattle, 1990, 2000, and
2010…………………………………………………………………………………..41
2.5 The spatial distribution of census block groups with educational attainment above
or below the median educational attainment in Seattle, 1990, 2000, and
2010………………………………………………………………………………….42
2.6 Difference in number of parks, above vs. below median household
income………………………………………………………………………………..49
2.7 Difference in areas (natural log of acres) of parks, above vs. below-median
household income……………………………………………………………………50
3.1 Example transition matrix……………………….……………………………….72
3.2 The spatial distribution of census block groups categorized as being greater than
or less than or equal to the city of Seattle median for median household income from
1990 –
2010…………………………………………………………………………………..77
3.3 The spatial distribution of census block groups categorized as being greater than
or less than or equal to the city of Seattle median for median home value from 1990 –
2010…………………………………………………………………………………..78
3.4 The spatial distribution of census block groups categorized as being greater than
or less than or equal to the city of Seattle median for educational attainment from
1990 –
2010..…………………………………………………………………………………79
LIST OF FIGURES (Continued)
Figure
Page
3.5 The graphical representation of the transition matrix showing the locations of
each census block group above or below the city median for each socioeconomic
category from 1990 to
2010.…………………………………………………………………………………80
3.6 The spatial distribution of socioeconomic change and gentrification in Seattle
from 1990 – 2000.……………………………………………………………………83
3.7 The spatial distribution of socioeconomic change and gentrification from 2000 to
2010..…………………………………………………………………………………85
3.8 Change in the African American racial composition in the Central Area from
1990 to 2000...………………….……………………………………………………87
3.9 Change in the African American racial composition in the Rainier Valley from
1990 to 2000...………………….……………………………………………………88
3.10 Change in racial composition of the Asian racial group in the University District
from 1990 to
2000………….………………………………………………………………………89
3.11 Change in the African American racial composition in gentrified census block
groups in the Central Area from 2000 to 2010……………………………….……...90
3.12 Positive change in African American population in gentrified census block
groups in various neighborhoods in Seattle from 2000 to 2010.
………………………………………………………………………………………..92
3.13 Spatial distribution of the greatest negative change in Asian racial groups within
gentrified census block groups from 2000 –
2010.………………………………………………………………………………….93
3.14 Greatest positive change in Asian population in a gentrified census block group
in the Downtown and Cascade neighborhoods from 2000 to 2010……………...…..94
3.15 Percent African American in 1990……………………………………………..97
3.16 Gentrified and declining census block groups, change in percentage of African
American residents within each census block group, and places where gentrification
was perceived………………………………………………………………………...98
LIST OF FIGURES (Continued)
Figure
Page
3.17 Expanded views of selected areas shown in Figure 16………………………100
4.1 Percent of census block groups in this category, in which there was some
(nonzero) investment in parks from (a) 1990-2000 and (b) 2000-2010……………122
4.2 Average size of park investment for census block groups which gentrified, did not
change or declined from (a) 1990-2000 and (b) 2000-2010……………………..…123
4.3 1990-2010 Change in the number of recreation parks within 800m of census block
groups in relation to gentrified and declined census block groups…………………125
4.4 1990-2010 Change in acreage of recreation parks within 800m of census block
groups in relation to gentrified and declined census block groups………………....126
4.5 1990-2010 Change in the number of natural passive parks within 800m of census
block groups in relation to gentrified and declined census block
groups…………………………………………………………………………….…128
4.6 1990-2010 Change in acreage of natural passive parks within 800m of census
block groups in relation to gentrified and declined census block
groups……………….................................................................................................129
4.7 1990-2010 Change in the number of multiuse parks within 800m of census block
groups in relation to gentrified and declined census block groups…………………131
4.8 1990-2010 Change in acreage of multiuse parks within 800m of census block
groups in relation to gentrified and declined census block groups………………....132
4.9 1990-2010 Recreation park investment compared to change in number of
recreation parks within 800m of census block groups in relation to gentrified and
declined census block groups………………………………………………………133
4.10 1990-2010 Recreation park investment compared to change in recreation park
acreage within 800m of census block groups in relation to gentrified and declined
census block groups………………………………………………………………...134
4.11 1990-2010 Natural passive park investment compared to change in natural
passive park acreage within 800m of census block groups in relation to gentrified and
declined census block
groups…………………………………………………………………………….....136
LIST OF FIGURES (Continued)
Figure
Page
4.12 1990-2010 Natural passive park investment compared to change in number of
natural passive parks within 800m of census block groups in relation to gentrified and
declined census block groups……………………………………………….………137
4.13 1990-2010 Multiuse park investment compared to change in multiuse park
acreage within 800m of census block groups in relation to gentrified and declined
census block
groups…………………………………………………………………………….....139
4.14 1990-2010 Multiuse park investment compared to change in number of multiuse
parks within 800m of census block groups in relation to gentrified and declined
census block groups……………………………………………….……………..…140
5.1 The modified Society and Land use
model…………………………..…………………………………………………...145
LIST OF TABLES
Table
Page
2.1 Seattle park classification………………………....……………………………..24
2.2 Socioeconomic variables………………………………………………………...33
2.3 City of Seattle median by socioeconomic variable, 1990 – 2010…………..........34
2.4 Numbers and area of parks by type in Seattle in 1990, 2000, 2010 and park
acreage added by decade and park type……………………………………………...37
2.5 The percentage of census block groups categorized as greater than or less than the
city of Seattle median for each socioeconomic characteristic by
year…………………………………………………………………………………...39
2.6 Mean number of recreation parks within 800m of census block groups above or
below the Seattle median for each socioeconomic category 1990 –
2010…………………………………………………………………………………..45
2.7 Group means for recreation park area within 800m of census block groups above
or below Seattle median for each socioeconomic category 1990-2010……………...45
2.8 Mean number of natural passive parks within 800m of census block groups above
or below the Seattle median for each socioeconomic category 1990 –
2010…………………………………………………………………………………..46
2.9 Group means for natural passive park area within 800m of census block groups
above or below Seattle median for each socioeconomic category 19902010…………………………………………………………………………………..47
2.10 Mean number of multiuse parks within 800m of census block groups above or
below the Seattle median for each socioeconomic category 1990 –
2010…………………………………………………………………………………..48
2.11 Group means for multiuse park area within 800m of census block groups above
or below Seattle median for each socioeconomic category 19902010…………………………………………………………………………………..48
2.12 Mean percentage of African Americans in census block groups above or below
the median for each socioeconomic variable, 1990 –
2010…………………………………………………………………………………..51
LIST OF TABLES (Continued)
Table
Page
2.13 Mean percentage of Asians in census block groups above or below the median
for each socioeconomic variable, 1990 –
2010…………………………………………………………………………………..51
3.1 Socioeconomic variables collected by the US Census Bureau (Census Bureau,
1990, 2000, 2010). These variables were used to analyze the process of
socioeconomic change from 1990 –
2010………………………………………………………………………………….70
3.2 City of Seattle overall median for selected socioeconomic characteristics for each
decennial census year (1990, 2000, and
2010)………………………..………………………………………………………..71
3.3 The transition trajectory for census block groups above or below the city median
in for each census year………………………….……………………………………71
3.4 The percentage of census block groups categorized as greater than or less than the
city of Seattle median for each socioeconomic characteristic by year………………74
3.5 The percentage of census block groups belonging to each transition for each
socioeconomic characteristic from 1990 to 2010…………………………………....76
4.1 Socioeconomic variables collected by the US Census Bureau (Census Bureau,
1990, 2000, 2010). These variables were used to analyze the process of
socioeconomic change from 1990 – 2010…………………………………….……118
1
Chapter 1 - Introduction
This dissertation assesses changes in socioeconomic variables and parks in
Seattle, Washington from 1990-2010 to explore causes of gentrification and possible
approaches to avoid the negative effects of gentrification. This study tests how the
built environment, more specifically urban parks, shape the communities around
them. The objectives of this study are:
(1) To gain a better understanding of socioeconomic and racial factors that are
associated with gentrification in Seattle, Washington,
(2) To develop a transferrable approach for assessing trajectories of
gentrification in urban areas, and
(3) To determine how the spatial distribution of different park types and
community investment in parks are associated with gentrification
The process of gentrification is a complex social, cultural, political, and
environmental phenomenon. Therefore, the overall research goal is to identify factors
that influence the spatial patterns of socio-economic change Seattle. The study is
structured around three questions:
(1) What are the spatial trends of access to differing publicly owned park
types based on differing socioeconomic characteristics from 1990 to
2010?
(2) What is the spatial pattern of gentrification?
(3) How are changes in park access and investment in parks related to
trajectories of gentrification?
The dissertation consists of five chapters. Chapter 1 is an overview of the
literature that motivated this research, focusing on concepts of environmental justice
and how they are related to environmental amenities and parks. It presents the
definition of gentrification and reviews competing theories of what causes
gentrification, including changes in environmental amenities. The history of the
Seattle Parks System and the typology used to classify parks for this study are briefly
2
described. The theoretical framework used to conceptualize and interpret findings of
this research is presented.
Chapter 2 examines spatial trends of access to differing park types and how
these trends are related to educational attainment, home value, and income in Seattle,
Washington from 1990 to 2000 and 2000 to 2010. Two questions are addressed in
this chapter: (1) How were park types distributed in Seattle in 1990 and how have
investments in park expansion changed those distribution as of 2000 and 2010? (2)
How does park access of varying types differ among census block groups based on
educational attainment, home value and income? This chapter sets the stage for
examining how environmental justice issues are related to park access.
Chapter 3 presents and applies a method to determine the spatial pattern of
gentrification (research question 2). Gentrification is defined based on data from the
US Census at the census block group level, and GIS techniques used to define and
detect gentrification are described. The resulting maps of gentrification are compared
to results from a media content analysis of perceptions of gentrification.
Chapter 4 introduces the concept of environmental gentrification and
addresses the third research question: How are changes in park access and investment
in parks related to trajectories of gentrification? Relationships among these variables
are assessed visually from maps based on GIS overlays. The findings are interpreted
to provide insight as to how communities may avoid certain trajectories of
gentrification based on community development, cooperation, and investment in
parks.
Chapter 5 is an overall conclusion and recommendations for future research.
1. Conceptual Framework
The Society and Land Use model (Platt, 2004) provides a framework for
understanding the iterative processes that shape the evolution of cities (Figure 4,
adapted from Platt). It depicts the dynamic interactions between the human/cultural
landscape, the environmental/physical landscape and the legal realm.
This model provides a framework to examine the dynamic feedback of processes
(e.g., gentrification and environmental gentrification) that are influenced by changes
3
in the physical landscape (e.g., increase or decrease in environmental amenities such
as parks), which are in response to or influence changes in the human/cultural
landscape (e.g. socioeconomic variables/ parks and land use systems) and the legal
realm (e.g. bonds/levies/community directed investment).
This model will help to frame this research because it allows for the
characterization of how policies/ park planning affects people and how people
inherently feed back into the process.
Figure 1. The society and land use model (adapted from Platt, 2004).
The circles or data sets are linked with vectors. The resource management
vector represents the process by which society organizes the use of land, water, and
air (or parks). This vector is representative of both the private and public authorities
that together determine how land is used or for example, what types of parks are
4
created in the landscape and where. The two input vectors (environmental perception
and economic and social data) that lead to the legal realm inform decision makers and
influence the output of resource-management. The environmental perception vector
refers to information regarding the state/condition of the physical environment, as it is
modified by human activities (the environmental impacts vector). In this study the
state or condition of the physical environment is characterized by the changes in the
amount of park area or the number of parks. The economic and social data vector
represents feedback on the financial and social effects of a land use, which helps to
address question one of this research. For example, the financial capability for a
community to influence the type of parks within the landscape, thus influence their
access may be related to their differing social and demographic standing compared to
other communities. The changes that occur over time to the economic and social
data vector help to address question two of this research. On the other hand the
decisions from the resource management vector that inform what types of parks are
associated within certain areas may have an influence on the social and economic
vector that feeds back into the legal realm that feed back into the resource
management vector from which decisions are made on land use systems, addresses
question three of my research. (Platt, p54).
2. Environmental Justice
This research explores environmental justice as it affects where people live
and how they interact with parks in a large urban area. The Environmental Protection
Agency (EPA) describes Environmental Justice as the “fair treatment and meaningful
involvement of all people regardless of race, color, national origin, or income with
respect to development, implementation, and enforcement of environmental laws,
regulations, and policies” (EPA 2001). Executive Order 12898 was enacted in 1994
and signed by President Clinton, which requires that all federal agencies work toward
resolving environmental justice issues by identifying and addressing
disproportionately high and adverse human health of environmental effects on people
with special attention to minority and low-income populations (EPA 1994).
5
Environmental justice activism began as a grassroots movement in the early
1980s. However, the concepts have been part of the politics of communities of color
for more than a century (Taylor 2009). A century ago, environmental activism was
marked by people of color banding together to fight for issues including, but not
limited to, improvements in housing conditions among slaves, opposition to
appropriation of land, erosion of treaty rights, and the share-cropping system (Taylor
2009). These are systematic forms of inequality that inform us about the treatment of
the other, based on ideals of race, class, and gender (Anderson and Collins 2013).
The environmental justice movement and body of literature thus stem from these
experiences and outcomes faced by people of color and make the link between racial
and economic inequalities and exposure to environmental hazards.
Studies in the 1980s linked race to exposure to environmental hazards. Two
early studies linked demographic patterns to commercial hazardous waste facilities
and uncontrolled toxic sites (described in United Church of Christ Commission for
Racial Justice 1987). One of these studies found that race was the most significant
variable to explain the location of hazardous waste facilities, and income and home
values were lower in communities with hazardous facilities compared to those
without hazardous facilities. The second report described a concentration of
uncontrolled toxic waste sites in Black and Hispanic neighborhoods, particularly in
urban areas (United Church Christ Commission 1987). In response to these findings,
a number of studies attempted to determine whether the coincidence of toxic waste
dumps and minority-dominated neighborhoods was the result of conscious
discrimination against vulnerable communities.
One longitudinal analysis of environmental justice revealed that the
percentage of African Americans in a census tract was a strong predictor of that tract
having a hazardous waste site, but did not show that hazardous sites were being
deliberately placed in Black communities. It also showed that working and lowermiddle income class neighborhoods had a higher chance of being disproportionately
chosen to host hazardous sites (Been and Gupta 1997). This research indicated that
environmental justice was not simply a racial issue, but social class was involved as
well. For example, people in manufacturing occupations were significantly
6
associated with hazardous facilities in census tracts (Anderton et al 1994). These
studies begin to reveal different effects of race, ethnicity, and class on the siting
processes of locally undesirable land uses. The environmental justice literature
examines these interrelated issues and how they work together as a system of power
and inequality that differentially affects groups depending on their social location
(Taylor 2009, Anderson and Collins 2013).
Environmental racism/justice through a systemic lens suggests that while
individuals and institutions may not be overtly seeking to create barriers, governance
may benefit some at the expense of others (Bolin et al 2002). "White privilege" is a
hegemonic form of racism that does not involve forms of hostile or discriminatory
acts typically associated with racism, but instead refers to unearned privileges and
benefits that are based on an individual’s degree of whiteness (Pulido 2000).
Fundamental to the concept of white privilege is the concept of social hierarchy that
incorporates cultural and social capital which creates an opportunity for resource
mobilization that is solely based on their degree of whiteness. White privilege and
processes like it are deeply imbedded in a legacy of racist ideologies, and even if they
are not intentional, they perpetuate spatial patterns in which people of color are
disproportionately exposed to environmental hazards. Protection from environmental
injustice, such as the distribution of hazardous sites, is also linked to the social
advantages of being part of a decision-making process, whereas non-whites tend to
lack adequate representation in policy and decision making processes (ShraderFrechette 2002; Agyeman 2005). Rather than invoking questions of environmental
racism, these studies appeal to the understanding of local historicized processes of the
social structure/institutions, and their racist underpinning that contribute to spatial
distribution patterns.
Many studies have sought to understand the historical, place-specific, and
institutional structures that create environmental inequalities. Discrimination in
housing and occupational opportunities restricted jobs and neighborhoods mainly to
whites around the Wagner Electric Plant in St. Louis; in the 1960s, blacks were
encouraged to move into the area and encountered a toxic brownfield (Hurley 1997).
Historical accounts reveal legacies of discrimination and racism linked to
7
environmental injustices. For instance, persistent spatial patterns of segregation led
to present conditions of whites living closer than blacks to toxic industrial sites in
Baltimore, Maryland (Boone et al 2009). This and other studies underscore the
importance of historical legacies, which may influence environmental justice (Krieg
2005, Downey 2005, and Saha and Mohai 2005).
Environmental injustice is not limited to exposure to hazardous facilities, but
can also involve exposure to other hazards. In Phoenix, lower socioeconomic and
ethnic minority groups were more likely to live in neighborhoods with more exposure
to heat stress and less shade and green open space to mitigate extreme conditions
(Harlen et al 2006). Childhood asthma hospitalization rates in New York City were
related to low median household income, high percentage of minorities, public
housing, and multiple environmental pollution burdens (Corburn et al., 2006). These
environmental justice studies made a conceptual distinction between the presence of
environmental hazards and the human health outcome/consequences of the exposure
(Harlen et al 2006). Highlighting the connections between environmental hazards
and human health outcomes provides a natural link to the injustices related to spatial
patterns of environmental amenities.
Recent research in environmental justice has addressed environmental amenities.
The term amenity is derived from the Latin word amoenus, meaning pleasant. An
amenity is a desirable or useful feature, facility, service or resource. The terms
"desirable" and "pleasant" imply a sense of value attributed to a feature, service, or
resource. Environmental amenities thus are desirable or pleasant features that are
built or occur naturally, making the area more desirable. Environmental amenities
can include air and water quality, percentage of pervious surface coverage, absence of
industry, or proximity to waste sites, type and amount of view (of water, forest, etc.),
access to parks or green space, or tree cover (Arnold and Gibbons 1996; Sander and
Haight 2012; Boone et al. 2009).
Park spaces are amenities because they provide economic, social, health and
ecological benefits. Urban parks are gateways to the natural world and imply a safe
city, where people are able to congregate and share their experiences (Burgess et al.,
1988). Exposure to nature through parks produces physiological responses, including
8
lower blood pressure and reduced muscle tension (Urlich 1981). Natural
environments with vegetation and water reduce stress in observers compared to urban
environments lacking these features (Schroader 1991). Environmental amenities also
affect property values, which increase near parks (Crompton 2001). The presence of
open space within 1500 ft raised housing property values by more than $2,600
(Bolitzer and Netusil 2000).
Researchers have identified a number of methods for valuing amenities (Sander
and Haight 2012). The production function method views the amenity as an input
into the production of a marketed good and its value is estimated from the price of
that good (e.g., Barbier 2007; Barbier and Strand 1998; Bell 1997; Klemick 2011;
Richmond et al. 2007; Sathirathai and Barbier 2001; Simonit and Perrings 2011).
Preference approaches use survey results to determine an individual’s willingness to
pay for an increase in a service or their willingness to accept compensation for a
decrease in a service (e.g., Boyle et al. 1994; Campos et al. 2009; Carson et al. 1992;
delSaz-Salazar and Rausell-Köster 2008; Earnhart 2006; McGonagle and Swallow
2005; Sayadi et al. 2009). The replacement cost analysis method uses the price of the
least-cost alternative means for providing a service or amenity as a proxy for its value
(e.g., Allsopp et al. 2011; Ashendorff et al. 1997; Hougner et al. 2006; Kovacs et al.
2011; NRC 2000). The levels of investment in maintenance of an amenity, such as a
park, may indicate the value people place on that amenity; this approach is used in
this study.
The spatial distribution of environmental amenities also produces inequalities
based on differences in access. A geographic approach to valuing environmental
amenities is to simply count or measure the number and area of certain types of
features, such as parks, that occur within a given distance of where people live.
Furthermore, the spatial distribution of amenities, such as parks, may have deep roots
based on historical racism, environmental injustice, and gentrification. This research
explores some of these connections.
9
3. Trajectories of Gentrification
This research is also motivated by controversy over the definition of gentrification
and its causes and consequences. Gentrification is described as “the simultaneous
phenomena of a physical, economic, social, and cultural change of a city” (Hamnett
1984, p286). Higher income groups move in to neighborhoods previously occupied
by lower-income groups, changing the physical, cultural and economic characteristics
of the neighborhoods. In US cities, gentrification literature began in the early 1950s
and grew rapidly throughout the 1960s. By the 1970s gentrification was evident in
many American cities (Smith 1987). As the number of gentrified cities grew so did
the literature on the topic.
The early literature on gentrification focused on socioeconomic and cultural
effects, displacement and the federal role in redevelopment, benefits of gentrification,
and the creation and destruction of communities. Little research addressed the causes
and the historical explanations of gentrification. However, over the last 30 years,
many scholars have contributed to theories about the causes, explanations and
significance of gentrification (Hamnett 1991). Early explanations of gentrification
were based on cultural or economic processes (Smith 1987).
3.1 Cultural theories of Gentrification
Lipton (1977) suggested that gentrification was caused by cultural factors. These
included a change in lifestyle and declining preferences for life in suburbia, later
marriage, fewer children, and younger homebuyers and renters who abandoned the
suburban life and redefined themselves in terms of city living. Another cultural
explanation for gentrification is that young people, or the “creative class” (Florida
2005) seek to establish socially distinct communities that allow for self-expression
and individualism (Winters 1978).
Ley (1978), argued that urban development was less about growth and more about
the quality of life for an individual. Drawing from Daniel Bell on post-industrial
society and Habermas on advanced capitalism, Ley suggested that ideals of
consumption rather than production were dominating patterns of society (Bell 1973,
Habermas 1973, Ley 1980). Ley (1980) described gentrification in terms of three
10
pillars: economy, politics, and culture. Post-industrial society brought with it an
economy that was transformed from production of goods to production of services, an
increased role of politics in decision-making, and a shift from a focus on individuality
to a focus on aesthetics. These three pillars form an emerging class of people (the
"gentrifiers"), who form what Veblen (1899) calls the leisure class, and whose
lifestyle is driven by status and consumption (Ley 1980). Ley’s (1980) theory helped
predict the type of city in which gentrification is likely to occur. However, the theory
did not attempt to explain patterns within cities, and it has been criticized for its
assumption that gentrifiers determine the urban landscape (Hamnett 1991).
3.2 Economic theories of Gentrification
Gentrification has also been linked to economic factors (Smith 1979). It became
more cost effective and profitable to rehabilitate an old neglected home within the
city limits than it was to buy a newly constructed residence in the suburbs. Coupled
with consumer preference to reduce commuting costs, this provides an explanation
for gentrification in economic terms. However, producers as well as consumers of
housing influence gentrification (Smith 1979 p.540). Smith (1979) argued that, “To
explain gentrification according to the gentrifier’s actions alone, while ignoring the
role of builders, developers, landlords, mortgage lenders, government agencies, real
estate agents and tenants is excessively narrow." The "rent gap" theory is based on a
related concept: that low rents in neglected neighborhoods provide an opportunity for
developers. Smith identifies the gap as the area between the potential rent and the
capitalized rent. The potential rent is what could be earned if the building in question
was being used at its highest and best use whereas the capitalized rent is the actual
rent (Smith (1979). In short the gap grows larger as the conditions of the home and
investment for repairs decline. Smith’s theory is well known however, there are
criticisms to his theory. One main criticism about this theory is that Smith is firm in
the idea that gentrification is production based and market driven and discounts
changes in demographics and culture as “surface froth” (Hamnett 1991). Many
researchers have challenged this idea and found that “gentrification is unable to exist
on its own without the gentrifiers (Hamnett 1991). Researchers like Mullins (1982)
11
and Beauregard (1986) explain that gentrification is complex process that must take
into account the consumption role of gentrifiers as well as cultural and demographic
changes in areas. For instance, they suggest that occupational changes from industrial
labor to more professional and managerial workforce are where the gentrifiers exist
(Hamnett 1991).
Theories of cultural, economic, and political effects on gentrification are
explored in this study. The economic and cultural conditions of a neighborhood may
create the potential for gentrification, while politics may influence the process of
environmental gentrification. For example, Ley’s cultural and economic pillars
coincide to form the gentrifiers, whose focus on aesthetics may increase
environmental amenities.
3.3 Environmental Gentrification
A primary motivation for this research is to explore how changes in the
environment are related to gentrification. The term "environmental gentrification"
was coined by Sieg et al. (2004) who showed that environmental policies aimed to
improve the quality of life for residents appeared to only benefit higher income
households, because the increases in the housing prices were not fully offset by the
air quality improvements that lower income residents can afford to enjoy. Low
income households tend to lose out when policies improve environmental quality and
initiate the process of environmental gentrification, where wealthy residents move to
communities with previously low levels of environmental quality. Housing prices
increase, and property owners rather than renters benefit from the policy change that
influenced the environmental improvement.
Banzhaf and McMormick (2007) suggested that the concept of environmental
gentrification was rooted in Tiebout’s notion of local public goods theory, which
provides an indirect explanation for the correlation between minorities and
environmental disamenities. Tiebout’s theory suggests that households will sort
themselves in different communities based on demands for amenities in the area such
as schools or environmental quality (Banzhaf and Walsh 2004). Banzhaf (2007)
suggested that, “If we consider environmental quality a normal good, even if all
12
people have the same tastes, we should expect that lower income households will
accept lower environmental quality in exchange for lower housing costs.” Because
minority households are typically also low-income they will be associated with lower
environmental quality due to the lower housing costs associated with those areas.
Research is needed to explore the connection between spatial patterns of
environmental amenities and processes of socioeconomic and racial change. This
study addressed environmental gentrification by looking at how investments in parks
are related to changes in socioeconomic and racial compositions of neighborhoods.
4. Park Planning in Seattle
The process of public park planning influences the spatial distribution of parks
and park investment, which in turn may influence environmental justice and
contribute to environmental gentrification. The Seattle Department of Parks and
Recreation acquires, develops, and maintains parks, recreation programs and other
projects with money from levies, donations, nonprofit organizations, grants, bond
measures and other funding sources. The process of park planning, implementation,
and management is done so primarily by the city however, the backbone of the
process is primarily the Seattle citizens. All bonds and levies must be approved by
voters, thus the success of the city park is in part due to the valuation that is placed on
the park system by the residents of the City Seattle. The creation of Seattle’s park
and playground system began in 1884 with a gift of property donated by David T.
Denny and his wife. From 1884 to 1904, acquisition of land and creation of new
parks was funded through appropriations by the city council and the Freeholders
Charter Revenue from fines, fees, and annual tax levies.
In the early 1900s, the Seattle Department of Parks and Recreation received
funding through park bonds. Park bonds issued in 1906 ($500,000).and two
subsequent bonds issued in 1908 and 1910 ($3 million) were used for the acquisition
and development of the Olmstead brother’s park plans. During this time 26 parks and
playgrounds were acquired and set the stage for what would be known as a worldclass park system. Throughout the 1920s and 30s more park bonds were issued
(totaling $889,000), and funds were devoted to development and improvements of
13
parks. In addition, funds were provided by the Emergency Unemployed Act during
the Great Depression and federal aid which provided funding for the development of
various recreation sites including: West Seattle Golf, Camp Long, and stadiums. In
the 1940s, funding for recreation programs, development and improvements was
provided from the federal Lanham Act ($70,000), a state grant ($1,225,000), and a
park bond ($2,500,000). Funding in the 1950s was low, because three separate bond
issues were defeated by Seattle voters. However, funding rebounded in the 1960s
with a bond approval for $4.5 million, a matching funds initiative, and a large park
bond for $65 million for acquisition and development of park land. Additional aid
was provided through the Forward Thrust program, a bond measure that was passed
in 1968 by Seattle and King County voters. This was a capital improvement bond
that honed on recreational improvements, acquisition of land and construction of park
facilities. During the 1980s, park funding was obtained through a renovation bond
issue, and the King County Open Space and Trails bond provided $41million that was
used to purchase and preserve more than 600 acres of land.
Funding for parks in the 1990s was obtained through the Seattle Center
Community Center Levy in 1991 ($48.9 million) and 1999 ($72 million). In 1999,
Seattle voters approved the renewal of the 1991 levy, renewing Seattle Parks and
Recreation’s commitment to renovate and expand old facilities and provide new
community centers to underserved areas within the Belltown, Highpoint, International
District/Chinatown, Jefferson, Northgate, Sand Point, Southwest, Van Asselt, and
Yesler neighborhoods. In 2000 an eight year Pro Parks Levy was approved for
$198.2 million for more than 100 different projects including development,
acquisition of land, maintenance, recreation programs, and environmental education.
Seattle voter’s excitement about Seattle’s public park system was reflected in the
passing of the Parks and Green Spaces Levy in 2008. This six-year levy provided
$148 million for up to 27 park acquisition projects, more than 60 park improvement
projects, and "neighborhood matching" funds for projects that are identified by
community groups throughout Seattle neighborhoods (Seattle Parks and Recreation)
This dissertation is relevant to parks and planning because it uses a park
typology to explore how access to park types varied by socioeconomic status and
14
race, and how park access varied over time. Information on neighborhood matching
fund investments in parks was analyzed to determine how investment is distributed
throughout the Seattle park system.
5. Park Typology
This study defined parks based on their amenities, use (active or passive), and
development level of the park. Active-use is characterized by the formal playing
fields, outdoor or indoor courts, swimming pools, community centers, restroom
facilities and/or picnic shelters that are available on site. Passive-use is defined as
having no formal fields for active play. They tend to have benches, or viewing areas,
and trails for walking, biking or hiking. The typology is adapted from King County
parks classification system and includes three park types: recreation, natural passive
and multiuse parks developed by Weems (2012). Recreation parks are highly
developed with formal facilities promoting active play such as play equipment, ball
fields, tennis courts, and community centers. These park types often lack an
abundance of greenspace, trees and other natural amenities that may be found in other
types of parks such as natural passive or multiuse parks. Natural passive parks are
characterized by their passive use nature and have no formal fields for active play.
They tend to have beaches, viewing areas, and trails for walking, biking or hiking.
Multiuse parks are a combination of both recreation and natural passive parks and
promote both active and passive use of the area. Differentiating between parks allows
for a better understanding of actual distribution and access to amenities within the
city.
6. Conclusion:
This research seeks to understand how park access has changed over time and how
that access may or may not have an influence over the changes in the economic and
social characteristics of neighborhoods (ie: gentrification or environmental
gentrification). This research utilizes a modified version of Platt’s Society and Land
Use model to help conceptualize the research. Chapter two addresses the access to
parks in Seattle over time and the relationship to socioeconomic variables.
15
7. References
Arnold, C.L., and C.J. Gibbons. 1996. Impervious surface coverage: The emergence
of a key environmental indicator. Journal of the American Planning Association. 62
(2): 243-58.
Agyeman, J. 2005. Sustainable communities and the challenge of environmental
justice. New York: New York University Press.
Andersen, Margaret L., and Hill Collins, Patricia. Race, Class, and Gender : An
Anthology. 8th ed. Belmont, CA: Wadsworth Cengage Learning, 2013.
Allsopp, M.H., W.J. de Lange, R. Veldtman. 2011. Valuing insect pollination
services with cost of replacement. PLoS One. 3: 1-8.
Anderton, D.L., A. B. Anderson, J.M. Oakes, and M.R. Fraser. 1994. Environmental
Equity: The Demographics of Dumping. Demography 31(2): 229-48.
Ashendorff C.L., M. A. Principe, A. Seely, J. LaDuca, L. Beckhardt, W. Faber, J.
Mantus. 1997. Watershed protection for New York City;s supply. Journal of the
American Water Works Association. 89: 75-88.
Banzhaf, H. and R. Walsh. 2004. Testing for environmental gentrification: Migratory
Responses to changes in environmental quality. Paper for AERE Workshop. Estes
Park, CO
Banzhaf, H., and E. McMormick. 2007. Moving beyond cleanup: Identifying the
cruicbles of environmental gentrification. Working paper 07-29, Andrew Young
School of Policy Studies.
Barbier, E.B. 2007. Valueing ecosystem services as productive inputs. Economic
Policy. 177-229.
Barbier, E.B., and I. Strand. 1998. Valuing mangrove-fishery linkages: a case study
of Campeche, Mexico. Environmental and Resource Economics. 12: 151-66.
Bell, D. 1973. The coming of post-industrial society: A venture in social forecasting.
New York, Basic books.
Bell, F.W. 1997. The economic value of saltwater marsh supporting marine
recreational fishing in the Southeastern United States. Ecological Economics. 21:
243-54
Been V. and F. Gupta. 1997. Coming to the nuisance of going to the barrios? A
longitudinal analysis of environmental justice claims. Ecology Law Quarterly 24: 156.
16
Berkes, F. 2002. Navigating Social-Ecological Systems. West Nyack, GB:
Cambridge University Press. Accessed March 9, 2016. ProQuest ebrary.
Bolin, B. A. Nelson, E. J. Hackett, K. D. Pijawka, C. S. Smith, D. Sicotte, E. K.
Sadalla, E. Matranga, and M. O’Donnell. 2002. The ecology of technological risk in a
Sunbelt city. Environment and Planning A 34:317–39.
Bolitzer, B. and N. R. Netusil 2000. The impact of open spaces on property values in
Portland Oregon. Journal of Environmental Management, 59 (3): 185-193.
Boone, C.G., G.B. Buckley, J.M. Grove, C. Sister. 2009.Parks and people: an
environmental justice inquiry in Baltimore, Maryland. Annals of the Association of
American Geographers. 99: 767-87.
Boyle, K.J., G.L. Poe, J.C. Bergstrom. 1994. What do we do about groundwater
values? Preliminary implications for a meta analysis of contingent-valuation studies.
American Journal of Agricultural Economics. 76: 1055-61.
Campos, P., J.L. Oviedo, A. Caparros, L. Huntsinger, I. Coelho. 2009. Contingent
valuation of woodland-owner private amenities in Spain, Portugal, and California.
Rangeland Ecology and Management. 62: 240-52.
Carson, R.T., R.C. Mitchell, W.M. Hanemann, R.J. Kopp, S. Presser, P.A. Ruud.
1992. A contingent valuation study of lost passive use values resulting from the
Exxon Valdez Oil Spill. Report to the Attorney General of the State of Alaska.
Corburn, J. J. Osleeb, and M. Porter. 2006. Urban asthma and the neighbourhood
environment in New York City. Health & Place 12:167-179.
delSaz-Salazar, S. P. Rausel-Koster. 2008. The double-hurdle model of urban green
areas valuation: dealing with zero responses. Landscape and Urban Planning.84:
241-51.
Downey, L. 2005. The unintended significance of race: Environmental racial
inequality in Detroit. Social Forces. 83 971-1007.
Earnhart, D. 2006. Using contingent-pricing analysis to value open space and its
duration at resitdential locations. Land Economics. 82: 17-35.
EPA. 1994. Executive Order 12898: Federal actions to address environmental justice
in minority populations and low-income populations. Federal Register 59: 2. Last
accessed 13 February 2016.
17
EPA 2001. Environmental Justice.
http://www3.epa.gov/environmentaljustice/index.html (last accessed 09 March 2016).
Habermas, J. 1973. Legitimation Crisis. (eng translation by T. McCarthy). Boston:
Beacon.
Hamnett, C. 1984. Gentrification and residential theory: a review and assessment, in
Herbert, D.R. and R.J. Johnston (eds) Geography and the urban environment.
Progress in research and applications, 6. John Wiley, London. p283 – 319.
Hamnett, C. 1991. The blind men and the elephant: The explanation of gentrification.
Transactions of the Institute of British Geographers. 16 (2): 173-89.
Harlen, S.L. A.J. Brazel, L.Prashada, W.L. Stefano, L. Larson. 2006. Neighborhood
microclimates and vulnerability to heat stress. Social Science and Medicine 63: 28472863.
Holling, C.S. 1973. Resilience and stability of ecological systems Annual Review of
Ecology and Systematics. 4:1-23.
Hougner, C., J. Colding, T. Soderqvist. 2006. Economic valuation of a seed dispersal
service in the Stockholm National Urban Park, Sweden. Ecological Economics. 59:
364-74.
Klemick, H. 2011. Shifting cultivation, forest, forest fallow, and externalities in
ecosystem services: evidence from the eastern Amazon. Journal of Environmental
Economics and Management. 61: 95-106.
Kovacs, K., T. Vaclavik, R.G. Haight, A. Pang, N.J. Cunniffe, C.A. Gilligan, R.K.
Meentemeyer. 2011. Predicting the economic costs and property values losses
attributed to sudden oak death damage in California (2010-2020). Journal of
Environmental Management. 92: 1292-1302.
Krieg, E.J. 2005. Race and environmental justice in Buffalo, NY: A ZIP code and
historical analysis of ecological hazards. Society & Natural Resources 18: 199-213.
Ley, D. 1978. Inner city resurgence units societal context, mimeo, paper presented to
the AAG Annual Conference, New Orleans.
_____ 1980. Liberal ideology and post-industrial city. Annals of the Association of
American Geographers. 70: 238-58.
Lipton, S. G. 1977. Evidence of central city revival. Journal of the American Institute
of Planners 43:136-47.
18
McGonagle M.P., S.K. Swallow, 2005. Open space and public access: a contingent
choice application to coastal preservation. Land Economics. 81: 477-495.
NRC. 2000. Watershed Management for a Potable Water Supply: Assessing the New
York City Strategy. National Academy Press, Washington, D.C.
Platt, R.H. 2004. Land use and society: geography, law, and public policy. Island
Press
Pulido 2000
Richmond, A. R.K. Kaufman, R.B. Myneni. 2007. Valuing ecosystem services: a
shadow price for net primary production. Ecological Economics. 64: 454-62.
Saha, R., and P. Mohai. 2005. Historical context and hazardous waste facility siting:
Understanding temporal patterns in Michigan. Social Problems. 52: 681-48.
Sathirathai, S. and E.B. Barbier. 2001. Valuing mangrove conservation, southern
Thailand. Contemporary Economic Policy. 19: 102-22.
Sander, H. and R.G. Haight. 2012. Estimating the economic value of cultural
ecosystem services in an urbanizing area using hedonic pricing. Journal of
Environmental Management. 113: 194-205.
Sayadi, S., M.C Gonzalez-Roa, J. Calatrava-Requena. 2009. Public preferences for
landscape features: the case of agricultural landscapes in mountainous Mediterrean
areas. Land Use Policy.26: 224-234.
Seattle, Parks and Recreation Department - Planning Documents 2009.
Schroeder, H.W., 1991. Preferences and meaning of arboretum landscapes:
combining quantitative and qualitative data. Journal of Environmental Psychology.
11: 231–248.
Simonit, S., C. Perrings. 2011. Sustainability and the value of the ‘regulating’
services: wetland and water quality in Lake Victoria. Ecological Economics. 70:
1189-99.
Shrader-Frechette, K.S. 2002. Environmental justice; Creating equality, reclaiming
democracy. Oxford, UK: Oxford University Press.
Smith, N. 1979. Toward a theory of gentrification; a back to the city movement by
capital, not people. Journal of American Planning Association. 45: 538-48.
Smith 1987
United Church of Christ Commission for Racial Justice. 1987. Toxic Wastes and
Race in the United States: National Report on the Racial and Socio-economic
19
Characteristics of community with hazardous waste sites. NY: Public Data Access,
Inquiries to the Commission.
Ulrich, R.S. 1981. Natural versus urban sciences: some pyscho-physiological effects.
Environment Behavior. 13: 523-56.
Taylor, D.E. 2009. The environmental and the people in American cities, 1600-1900s.
Duke University Press, Durham and London.
Veblen, T. 1899. Theory of the leisure class. Macmillan.
Walker, Brian, and Salt, David. 2006. Resilience Thinking : Sustaining Ecosystems
and People in a Changing World. Washington, DC, USA: Island Press. Accessed
March 8, 2016. ProQuest ebrary.
20
Chapter 2: Access to parks in Seattle: relationship of park type to socioeconomic
variables, 1990 to 2010
Abstract
Little is known about how environmental amenities, such as parks, interact with
socio-economic and racial characteristics of neighborhoods over time. This study
examined the relationships among public park system development over time, park
typologies, and socio-economic and racial patterns, in order to understand how park
system evolution is related to environmental justice. The spatial trends in access to
recreation, natural passive, and multiuse natural parks were determined and related to
educational attainment, home value, household income, and race for 481 census block
groups in Seattle, Washington from 1990 to 2010. Recreation parks are defined as
being highly developed with formal facilities promoting active play. Natural passive
parks are characterized by their passive use nature, viewing areas, and trails for
walking, biking or hiking. Multiuse parks are a combination of both recreation and
natural passive parks and promote both active and passive use of the area. Park
access is defined as the number of parks and amount of area within800m of a Census
block group. The study revealed that park access differed among socio-economic
groups and by park type. Significantly greater proportions of African Americans and
Asians were found in Census block groups with below-median levels of education,
income, and home value, which in turn had access to significantly larger numbers and
more area of recreation parks. In contrast, census block groups with above-median
income, home value and education and significantly fewer minorities had access to
greater numbers and more area of natural passive and multiuse natural parks. These
patterns did not change much over the study period (1990 to 2010). Park investment
had the effect of maintaining differences in park access based on socioeconomic
variables, even though new natural passive parks were added to areas formerly
lacking natural passive parks. Access to natural passive parks varied by income and
home value, but not by education level, implying that educational status and park
access are indicators of ongoing demographic changes in neighborhoods. This work
raises the question of whether investments in park improvements led or followed the
21
changes in educational attainment, home value, and income, suggesting that further
work is needed to test hypotheses of environmental gentrification in Seattle.
Keywords: environmental justice, environmental gentrification, parks, GIS
22
1. Introduction and Research Questions
Parks and other natural amenities enhance the quality of life in cities and are a
central value in modern society (Burgess, et al. 1988; Kornblum 1987). Parks offer
social, mental and physical health as well as economic and environmental benefits.
Parks promote social interaction and community development (Moore and Woodrow
1982). Access to parks improves mental health, reduces stress and promotes a sense
of tranquility (Urlich 1981; Kaplan 1983; Chiesura 2004, Hartig et al. 1991; Conway
2000; Mass et al. 2009). Parks in urban ecosystems provide increased environmental
health through air filtering, microclimate regulation and noise reduction (Bolund and
Hunhammar 1999). Property value increases with proximity to parks (Crompton
2001; Espey and Owusu-Edusei 2001; Lutzenhiser and Netusil 2001; Ernst and
Young 2003; Troy and Grove 2008; Voicu and Been 2008). Because of their social,
economic, and environmental benefits, parks are critical components of publicly
owned space in the US (Zhang et al. 2011; Talen 2010; Burgess et al 1988).
1.1. Park typologies
There are many types of parks, and differences among parks may affect how
benefits are expressed and who experiences them. Several different park typologies
have been proposed based on historical trends, characteristics of surrounding
neighborhoods, and other factors.
A historically based typology of parks described four major time periods that
coincided with a type of park developed to remedy the social problems of the time
(Cranz 1982). These four periods and associated park types were: the pleasure
ground (1820 – 1900), the reform park (1900-1930), the recreation facility 1930 –
1965, and the open space concept (1965 - ) (Cranz 1982). Pleasure ground parks,
such as those designed by Frederick Olmstead, were established around the edge of
the city, and while in theory these parks were created for everyone, in reality this park
type were often inaccessible to working class individuals. Therefore, beneficiaries of
"pleasure ground" parks were typically upper middle class populations. Reform parks
were established between 1900 and 1930, in response to the injustice associated with
23
the location of the pleasure ground parks and to benefit the wellbeing of the working
classes. Reform parks were small, and some featured play equipment for children.
Recreation facility parks were based on the establishment of a nationwide effort to
extend parks and playgrounds to areas that had not yet received them. The recreation
facility era coincided with the establishment of stadiums, parking lots, and asphalt
ball courts. During Open Space period, parks promoted free-form play and other
activities that would revitalize a city. The open space system helped to transform
how parks were used and created a network linked by open space (Cranz and Bolund,
2004).
An alternative park typology is based on the characteristics of the
neighborhoods that surround the park, including housing density and income level of
the neighborhood (Marcus et al 1998). Parks in areas of low housing density and low
to medium income neighborhoods have a typology similar to the recreational facility
model described by Cranz (1982). These parks are found in older neighborhoods
typically in the suburbs, were developed after World War II, and feature sports
facilities and playground equipment. They were established primarily for community
gatherings and formal programs.
A third park typology, used in this study, groups parks based on their use
(active or passive), and development level of the park. The typology is adapted from
King County parks classification system and includes three park types: recreation,
natural passive and multiuse parks developed by Weems (2012) (Table 1).
Recreation parks are highly developed with formal facilities promoting active play
such as play equipment, ball fields, tennis courts and community centers. These park
types often lack greenspace, trees and other natural amenities that occur in natural
passive or multiuse parks. Natural passive parks are characterized by their passive
use nature and have no formal fields for active play. They tend to have beaches,
forests, fields, viewing areas, and trails for walking, biking, or hiking. Multiuse parks
are a combination of both recreation and natural passive parks and promote both
active and passive use of the area. This park typology is useful for assessing the
amenities provided by parks.
24
Table 1. Park classification system based on park use (active or passive), and
development level of the park. (Source: Weems, 2012).
1.2. Evolution of park systems over time - the role of government
Federal and state governments have influenced the evolution of parks through
park funding, which favored the development of certain types of parks in certain
places. For example, the Federal Highway Administration's Recreational Trails
Program (RTP) provides funding to states to create and maintain recreational trails
and trail facilities for hiking, bicycling, inline skating, equestrian riding, all-terrain
vehicle riding, and other off-highway motorized vehicles (Federal Highway
Administration 2016). State governments have also influenced park systems through
certain types of funding. For example, the Washington State Recreation and
Conservation Office (RCO), established in 1964, provides grants (matching funds) to
state and local agencies, nonprofits, corporations, tribes and private landowners,
which can be used to acquire land; develop, restore and/or beautify wildlife habitat;
improve outdoor recreation properties; and protect farmland (RCO 2015). Since its
inception, the RCO provided $1.7 billion in grants for approximately 7500 projects,
and since 1990 it has provided about $60 million for approximately 230 projects
annually (RCO 2015). Based on the grant conditions it is plausible to think that the
RCO has played a significant role in influencing park development especially
regarding the types of parks. For instance restoration and beautification of wildlife
habitat may be associated with Natural passive type parks whereas; improvements to
outdoor recreation may be targeted to recreation parks.
1.3. Park access and environmental justice
Because park types vary among neighborhoods, park access, (defined as the
number of or the amount of area of each park type within 800m of a Census block
group) tends to vary among socioeconomic groups. Social factors affect may be
25
intertwined with ecological conditions of an area (Swyngedouw 1999) and may
produce inequities in park access. For example, more white park users of Chicago’s
largest park (Lincoln Park) came from neighborhoods nearest to the park, while many
Black, Latino, and Asian groups lived farther from the park (Gobster 2002). Lowincome areas in Los Angeles with high concentrations of poverty had lower levels of
park accessibility and resources (Wolch et al 2005). Education level was
significantly positively related to the percentages of green space and urban canopy
coverage in Central Indiana (Heynen and Lindsey 2003).
Access to parks and their benefits involves environmental justice, and differences
in access to parks among socio-economic groups may involve environmental
injustice. Environmental justice activism began as a grassroots movement in the
early 1980s, which recognized the connection between racial and economic
inequalities and exposure to environmental hazards. A number of studies suggested
that low income and minority residents were disproportionately exposed to
environmental hazards compared to communities with higher socioeconomic status
(e.g. United Church Christ Commission 1987; Been and Gupta 1997). Although the
environmental justice literature has primarily focused on environmental disamenties
or locally undesirable land uses, recent environmental justice research has expanded
to include environmental amenities such as parks.
1.4. Research questions: Park system evolution, park types, and environmental justice
This study examined the relationships among park system development over
time, park typologies, and socio-economic and racial patterns, in order to understand
how park system evolution is related to environmental justice. Little is known about
how park types interact with socio-economic and racial characteristics over time.
Researchers have often failed to differentiate between park types when considering
park distribution and park accessibility to different socioeconomic groups (e.g. Wolch
et al 2005, Boone 2009, Zhou and Kim 2013).
Taking advantage of available data on park size and type, as well as census
data from 1990 to 2010, this study examined the spatial trends in access to parks and
26
their relationship to educational attainment, home value, household income, and race
in Seattle, Washington from 1990 – 2010. This study addressed three questions:
1) How were various types of publicly owned parks distributed in Seattle in 1990, and
how have investments in the park system changed park distributions from 1990 to
2010?
2) How did access to parks of various types vary among census block groups by
educational attainment, home value, and income in Seattle in 1990, 2000, and 2010?
3) How were differences in park access related to racial groups in Seattle in 1990,
2000, and 2010?
1.5. Research hypotheses
Hypothesis 1.1: From 1990 to 2010, park area increased and new parks were
established in areas formerly lacking parks.
Hypothesis 1.2: Park investment reinforced unequal access to parks among
neighborhoods differing in education, income, and home value.
Parks are created by donation of land, land trading between city departments,
or purchasing of new areas. Multiuse and natural passive parks are often donations or
purchases of adjacent parcels of undeveloped land, which are available only in certain
areas. Recreation parks are also created through land donations, trading, and
purchasing of new areas. However in some cases, recreation parks are created when
schools close and the playground is converted into a recreation park. Creation of
these parks is often dependent on community involvement.
Hypothesis 2.1: Census block groups with higher income, educational attainment and
home value had better access to natural passive and multiuse parks compared to
census block groups with low income, educational attainment and home value.
Educational attainment, home value, and household income are correlated.
Individuals with the ability to pay for amenities such as more green space would be
27
expected to live in closer proximity to natural passive or multiuse parks compared to
those unable to afford such amenities. This is consistent with local public goods
theory, which suggests that households will sort themselves in different communities
based on demands for amenities in the area such as schools or environmental quality
(Banzhaf and Walsh 2004). Multiple studies have shown disparities in access to
parks based on socioeconomic status. For example, canopy cover and median
household income are positively correlated in Milwaukee, Wisconsin (Heynen et al.,
2006). Highly educated neighborhoods tended to have greater canopy
coverage/greenspace (Heynen and Lindsey 2003). In this study, natural passive and
multiuse parks have more green space than recreation parks, leading to the
expectation that they will be associated with high income areas.
Hypothesis 2.2: Census blocks with lower educational attainment, income, and home
value had more access to recreation parks compared to census block groups with high
income, educational attainment, and home value.
In a previous study on park access and crime in Seattle, found that recreation
parks occurred in neighborhoods with lower income, educational attainment and
home value (Weems 2012). Other studies have shown strong correlations between
park accessibility and income (Boone et al 2009).
Hypothesis 3: Census block groups with household income, home value and
educational attainment above the city median had overall higher park access to all
park types and lower proportions of minorities compared to census block groups
below the city median. These differences were expected to increase over time.
Research has repeatedly shown that race and socioeconomic conditions are
interrelated. Large gaps in education persist between minorities and white
populations (American Council on Education, 2004). Nonwhite children are more
likely to live in poverty than white children (Costello, Keeler, & Angold, 2001;
National Center for Education Statistics, 2007). Thus we can expect that minority
racial groups will be highly correlated with lower income, education, and, home
28
value, therefore if the opposite is correlated to more park access of all types we can
expect that minorities will have less access to parks.
2. Study site
Seattle, Washington is the largest city in the Pacific Northwest with an area of
217 square km and a population of 608,255 (US Census, 2010). The racial fabric of
the city is mostly Whites (69.5 percent), Asians (13.8 percent), and Blacks (7.9
percent). The remaining 8.8 percent includes American Indian and Alaska Native,
Native Hawaiian and other Pacific Islanders, and two or more races together. In
contrast to the state of Washington as a whole, Seattle is a highly educated city. As
of 2010, 57.4 percent of the population age 25 and older held a bachelor’s degree or
higher, compared to only 31.9 percent in the state of Washington and 28.8 percent in
the US (United States Census Bureau, 2009-2013). However, less than half of the
population of Seattle (46.8 percent) owns a home, compared to 63.2 percent in the
state of Washington and 64.9 percent in the US. The proportion of households below
the poverty line is 13.6 percent, which is comparable to the rest of the state of
Washington.
The selection of this study site was based on two key factors:
(1)
Seattle’s extensive change to its socioeconomic and racial history should
be explored in a spatial context. There is a great deal of media attention
surrounding this city and its underlying social processes. Seattle’s Central
District transformed from predominately working class African American
neighborhoods to mostly White, Asian American, and African American
professionals, via gentrification (McGee 2008). More recently, Seattle
was featured as being one of the “most gentrified cities in America”
(Business Insider.com). In February 2014, The Seattle Times reported that
“two masked men blocked a Microsoft employee shuttle on Capitol
Hill…condemning the neighborhood’s transformation into an ‘upscale
yuppie playground’…[with] skyrocketing rents that are displacing lessaffluent, longtime residents” (The Seattle Times 2014).
29
(2)
Seattle’s park system dates back to 1884, and consistent city investment
has made it leading park system in the world and an ideal place to explore
environmental gentrification. Seattle parks are funded in part from Seattle
residents, who have strongly supported park development. In recent years
voters three levies: the Community Center Levy ($36 million, 1999), the
Pro Parks Levy $198 million, 2000), and the Parks and Green Spaces Levy
($145 million, 2008). In addition, the Seattle Department of Parks and
Recreation also sought federal, state, and regional funding (Seattle Parks
and Recreation, 2014) and has defined a goal of accessibility, to ensure
that all residents live within 800m of a park (Harnik 2004).
30
Figure 1 Census block groups (n=481) in the city of Seattle as defined by the US
Census Bureau in 2010.
31
3. Data sources and Methods
This study used a combination of descriptive and parametric statistics to
examine spatial and temporal patterns of parks and socio-economic variables.
Descriptive statistics (nearest neighbor and directional distribution ellipses) were used
to describe and depict the spatial distributions of park types in Seattle in 1990, 2000,
and 2010. Median levels of educational attainment, home value, and household
income were determined for the city of Seattle for each of three years: 1990, 2000,
and 2010. Each census block group was identified as being above or below the
median value for education, home value, and household income in each year. Park
access was determined using a GIS-based buffer analysis that identified the number
and area of each park type (recreation, natural passive and multiuse parks) within
800m of each census block group boundary. Then, park access values were grouped
into above-median and below-median categories based on educational attainment,
home value, and household income, for each of three years: 1990, 2000, and 2010.
The relationship of park access (numbers and areas of each of three park types) to
socio-economic variables (education, home value, and income) and year (1990, 2000,
and 2010) was determined using a two-way Analysis of Variance (ANOVA) with the
post-hoc Tukey multiple comparisons test. Also, the relationship of minority racial
groups (the proportion of population in a census block group that were racial
minorities) to socioeconomic variables and year was tested using a two-way ANOVA
and Tukey test.
3.1 Socioeconomic data sources
This study used socioeconomic data for the 1990, 2000, and 2010 US census.
The data were acquired from the National Historical Geographic Information System
through the Minnesota population center, University of Minnesota.
The geographic unit of analysis for all census data was the census block
group. A block group is a statistical subset of a census tract, and generally contains
between 600 and 3,000 people (US Census Bureau 2012). Each block group is
uniquely identified allowing it to be represented graphically with a GIS. As of 2010,
Seattle had a total of 481 block groups (Figure 1). In 2000 there were 570 census
32
block groups, and in 1990 there were 630. Due to the changing boundaries of census
block groups over time, this study aggregated the 1990 and 2000 census block group
data to the 2010 census block group area based on weighted area. Over time the
Census Bureau changes the Census delineations for various reasons for example
changes in population size. When changes are made the census data no longer aligns
from census year to year. For this study it was important to establish one Census
Block group boundary in order to assess all of the data from changing years.
Three socioeconomic variables were selected for this study (Table 2).
Variables were based on previous research in the quality of life, environmental
justice, environmental gentrification and environmental justice literature (e.g., Jensen
2008, Sieg et al. 2004 and Boone 2009). Median household income and median
home value were used to measure the economic status of the population. The
percentage of residents 25 and older with bachelor degrees or higher was used as a
measure of education (Table 2). In addition, racial groups were considered, including
Whites and minorities (Blacks and Asians). It is important to note that many factors
may influence changes in the park access, however only four factors were used in this
study.
The socioeconomic census data in .csv files and census block groups in
shapefile format were downloaded from the National Historical Geographic
Information System website. Using ArcGIS10.2.2, the .csv files and shapefiles were
imported into the ArcMap data frame. Using the join function, the .csv files and the
shapefiles were joined together using their unique identifying object ID number,
allowing each census block group to receive its corresponding socioeconomic value.
33
Table 2. Socioeconomic variables collected by the US Census Bureau (Census
Bureau, 1990, 2000, 2010). These variables were used to analyze the process of
socioeconomic change from 1990 – 2010.
3.2. Park data sources
Multiple datasets on parks in Seattle were obtained from the Washington State
Geospatial Data Archive (WAGDA) (University of Washington 2007) and King
County GIS data portal (University of Washington 2007 and (King County GIS
2014). The spatial reference used for all park datasets was Washington State Plane
North; FIPS Zone 4601. The units were originally in feet but were converted to
meters for this analysis. The shapefiles provided location data as well as attribute
information including the parcel identification numbers (PIN), site names, owners,
and managers of each park.
Limitations with the park data included missing acquisition dates and missing
park parcels. To mitigate the issues the data were examined for inconsistencies, and
each dataset obtained from WAGDA and King County GIS was compared with the
others to ensure that all parks and their attributes were included. Information
regarding park investment and improvement, acquisition dates, historical facts and
other anecdotal information pertinent to the overall history of the Seattle park system
was obtained from Seattle parks and recreation documentation, Seattle city Archives,
and an interview with city of Seattle’s Property and Acquisition Manager.
The park shapefiles were imported into ArcGIS10.2.2. Using the editing
function, each park parcel was updated to include accurate acquisition dates,
investment amounts and other notable information.
34
3.3. Defining above- and below-median census block groups
The overall median household income, median home value, and educational
attainment were calculated for the City of Seattle for 1990, 2000, and 2010 (Table 3).
Each census block group was classified as above (labeled as A) or below (labeled as
B) the overall city median for each of these variables in each census decade.
Table 3. City of Seattle overall median for selected socioeconomic characteristics for
each decennial census year (1990, 2000, and 2010).
3.4. Park types, distribution and access
Each park was classified according to Weems (2012) into one of three
categories: recreation, natural passive, or multiuse. The spatial distribution of parks
in each year was quantified using the average nearest neighbor analysis to determine
if parks were clustered, evenly distributed, or randomly distributed. Nearest neighbor
analysis was conducted using ArcGIS10.2.2. The directional distribution (standard
deviation ellipse) tool in ArcGIS10.2.2 was used to calculate the mean center of all
features and identify if the park parcels by type were trending toward a particular
direction. The directional distribution tool creates an ellipse that is centered on the
mean center of a group of features. The direction represents the rotation of the long
axis measured clockwise from noon. A normal distribution with one standard
deviation, which was used in this study, encompasses 68 percent of the all feature
centroids (ESRI 2009). This process was used to measure the direction of all park
parcels in each of the three park types, for parks established prior to 1990 and park
parcels added between 1990 to 2000, and 2000 to 2010.
35
Access to parks was defined as (1) the number and (2) the area of parks
located within 800m of a census block group. Based on studies of park accessibility,
the standard distance that people are willing to walk to reach a park is roughly 400m
(Forsyth 2000, Nicholls 2001, Lindsey et al. 2001). However, in this study, an 800m
buffer in 100m increments was defined from the perimeter of each census block
group based on the stated goal of the Seattle Department of Parks and Recreation to
ensure that all residents have at least one park within 800m (Harnik 2004).
Researchers measuring access to parks within walking distance suggest using
network analyses based on entry points to parks (Liu and Zhu 2004; Zhou and Kim
2013). However, this approach was not feasible for this study because it was
impossible to identify specific points of entry for large parks within this dataset.
Instead, the number or area of parks of each type within a particular distance was
defined as a measure of access to parks, similar to Boone et al. (2009). Using
ArcGIS10.2.2, 800m distances from the perimeter of each census block group were
generated using the buffer tool set to side type FULL (which generated buffers
around the polygons and contains and overlaps the areas within the perimeter of each
polygon) and dissolve type NONE (which allowed individual buffers for each
polygon to be maintained, regardless of overlap). The number of parks within each
100m-increments (including the parks that may fall within the census block group) up
to 800m were counted for all census block groups and for each subset of census block
groups above and below the city median for educational attainment, home value, and
income.
Another measure of access to parks is the amount of park area within a certain
distance of a census block group. Using the buffer tool in Arc GIS 10.2.2, an 800m
buffer set to side type FULL (which generated buffers around the polygon and
contains and overlaps the areas within the perimeter of the polygon) and dissolve type
NONE (which allowed individual buffers for each polygon to be maintained,
regardless of overlap) in 100m increments was placed around the perimeter of each
census block group. Using the identity tool, each park type was overlaid on the
buffers to compute the geometric intersection of the park type and the buffer. The
output shapefile provided the area intersection of the two layers in acres.
36
3.5. Statistical analysis
The statistical significance of differences in park access by education, home
value, and income was tested using a two-way ANOVA followed by a post-hoc
Tukey’s multiple comparisons test using the R statistical program. The Two-way
ANOVA is used to analyze the difference and significance among group means. The
Tukey’s multiple comparisons test then identifies pairwise significant differences
within the dataset. Dependent variables in the ANOVA were (1) the number of parks
within 800m, and (2) the park area within 800m. Independent variables in the
ANOVA were census years (1990, 2000, and 2010) and above or below-median
groups with respect to education, home value, and income. Pairwise comparisons
with a p adjusted value < 0.05 were considered significant.
The statistical significance of racial difference in education, home value, and
income was tested using a two-way ANOVA followed by a post-hoc Tukey’s
multiple comparisons test using the R statistical program. Dependent variables in the
ANOVA were (1) percent African Americans, (2) percent Asians. Independent
variables in the ANOVA were census years (1990, 2000, 2010), and above or belowmedian groups with respect to education, home value, and income.
4. Results
4.1. Park distribution by type and changes in spatial distribution of parks over time
In 1990 there was a total of 268 parks in Seattle, covering 4,969.2 acres
(Table 3). Of these, 124 were recreation parks, 80 were natural passive parks, and 64
were multiuse parks (Table 4). In 2010, Seattle had 366 parks covering 5880.3 acres.
Of these, 164 (1,082.5 acres) were recreation parks, 126 (1,773.4 acres) were natural
passive parks, and 76 (3024.4 acres) were multiuse parks.
37
Table 4. Numbers and area of parks by type in Seattle in 1990, 2000, and 2010 and
park acreage added by decade and park type.
Results of the average nearest neighbor test indicate that recreation and natural
passive parks in Seattle were clustered in 2010 (p < 0.0001). However, multiuse
parks were randomly distributed.
Park parcels added to the Seattle Park System can be added to existing parks
or form brand new parks. From 1990 to 2000, 893 park parcels were added to the
Seattle Park System: 49 recreation parcels (51.7 acres), 801 natural passive parcels
(483.2 acres), and 43 multiuse parcels (139.1 acres). From 2000 to 2010, 283 park
parcels were added to the park system: 42 recreation parks (35.0 acres), 207 natural
passive parks (119.4 acres), and 34 multiuse parks (58.3 acres) (Table 4).
In 1990, the mean center of the directional distribution ellipse for recreational
parks encompassed center and portions of South Seattle. The ellipse for natural
passive parks trended toward northwest Seattle. Multiuse parks were concentrated in
the southeast sections of the city along the eastern coastline (Figure 2). From 1990 to
2000, parcels in natural passive parks were mostly added in the southern regions of
the city, whereas multiuse parks were added mostly in the northeast, and recreation
parks were added mostly to the center of the city (Figure 2 B). From 2000 to 2010,
parcels in natural passive parks were mostly added in the southern regions of the city,
whereas multiuse park parcels were added mostly in the northwest, and recreation
38
parks were mostly added in the center of the city (Figure 2 C). Between 1990 and
2010, natural passive and recreation park parcels were added in areas that lacked
these park types in 1990 (Figure 2). In contrast, new multiuse park parcels were
added to areas that already had this type of park in 1990 (Figure 2).
Figure 2. The directional distribution of different park types added to the
Seattle Park System 1990 to 2000 and 2000 to 2010. (a) all parks by type as of
1990, (b) parks added by type from 1990 to 2000, (c) parks added by type
from 2000 to 2010.
4.2. Locations of above- and below-median census blocks
From 1990 to 2010, the education level in Seattle increased from a median of
38 to 56% of people aged 25 yrs or older with a Bachelor's degree (Table 2). Over
the same period, median home value increased from $137,900 to $456,200, and
median household income increased from $29,353 to $60,655 (Table 2). The
proportions of census block groups that were above or below the median value did
not change significantly from 1990 to 2010 (Table 5). Spatial patterns of above- and
below-median census block groups changed only slightly from 1990 to 2010. In 1990
census block groups that were above the city median for household income were
39
mostly concentrated in north Seattle, and along the Puget Sound, and most of the
census block groups below the city median were found in south Seattle (Figure 3-A).
These patterns persisted in 2000 and 2010 (Figure 3-B and C).
Table 5. The percentage of census block groups categorized as greater than or less
than the city of Seattle median for each socioeconomic characteristic by year.
Census block groups that were above the city median for home value and
education also were mostly found along the shores of Puget Sound and Lake
Washington and in the mid northern sections of the city, and these patterns persisted
in 2000 and 2010 (Figure 4, Figure 5). The number of census block groups with
above-median educational attainment increased over time, but this increase was
concentrated in the northern and southwest sections of the city (Figure 5).
40
Figure 3. The spatial distribution of census block groups with household income
above or below the median household income in Seattle for 1990, 2000, and 2010.
41
Figure 4. The spatial distribution of census block groups with home value above or
below the median home value in Seattle for 1990, 2000, and 2010.
42
Figure 5. The spatial distribution of census block groups with educational attainment
above or below the median educational attainment in Seattle for 1990, 2000, and
2010.
43
4.3. Relationships among park area, park type, home value, income, and education
Visual comparisons of distributions reveal some relationships between park
types and socio-economic variables (Figures 2 to 5). In 1990, most multiuse and
natural passive parks appeared to be located in areas along the shorelines where
census block groups were above the overall Seattle median for household income,
home value, and educational attainment. However, recreation parks appeared to be
located in areas where household income, home value, and educational attainment
were below the city median (Figures 2-A, 3-A, 4-A, and 5-A).
From 1990 to 2000, recreation park parcels added to the Seattle park system
were primarily located in areas that were below the city median for household
income, home value, and educational attainment. Multiuse park parcels were
primarily added to areas along the shorelines and sections of south Seattle, in census
block groups above the city median for educational attainment and home value, but
below the Seattle median for household income. Natural passive park parcels were
primarily added in south Seattle, within narrow corridors that were above the city
median for household income and below the median for educational attainment and
home value (Figures 2-B, 3-B, 4-B, and 5-B).
From 2000 to 2010, recreation park parcels were mostly added near
downtown Seattle and extending into southern and northern Seattle. These new or
expanded recreation parks were located in census block groups that were above the
Seattle median for household income and educational attainment, but below the
median for home value. Natural passive parks were mostly added in the southeast
and southwest sections of the city in census block groups that were below the Seattle
median for household income, home value and educational attainment and also in
some narrow corridors of above-median census block groups. Multiuse parks were
added in the northwest section of the city and coincided with areas that were above
the city median for household income, home value and educational attainment
(Figures 2-C, 3-C, 4-C, and 5-C).
44
4.4. Access to Recreation Parks and change over time by education, home value, and
income
Statistical analysis of park access data confirms the relationships of socioeconomic variables to park type distributions and changes in park distributions over
time. Overall, the numbers of recreation parks was significantly higher (4 or 5 parks
vs. 3 parks) for census block groups that were below the median in household income
in 1990, 2000, and 2010. However, numbers of recreation parks did not differ
significantly among census block groups based on home value or education.
Numbers of recreation parks increased significantly from 1990 to 2000 for census
block groups that were below the median for education and from 2000 to 2010 for
census block groups that were below the median for home value and household
income (Table 5). In contrast, the number of recreation parks did not change from
1990 to 2000 for above-median census block groups, but it increased significantly
from 2000 to 2010 for census block groups that were above the median in education
and household income (Table 5).
The area of recreation parks was significantly higher (roughly twice) for
census block groups that were below the median in education in 1990, 2000, and
2010, and for those that were below the median for home value in 2000 and 2010.
However, the area of recreation parks did not differ significantly among census block
groups based on household income. There were no significant changes in recreation
park area based on education, home value, or household income from 1990 to 2010
(Table 6).
45
Table 6 Mean number of recreation parks within 800m of census blocks groups above
or below Seattle median for each socioeconomic category 1990 – 2010.
Table 7 Group means for recreation park area within 800m of census block groups
above or below Seattle median for each socioeconomic category 1990-2010.
4.5. Access to Natural Passive Parks and change over time by education, home value,
and income
Overall, the numbers of natural passive parks was significantly higher for (2
or 3 parks vs. 1 park) for census block groups that were above the city median for
education, home value and household income in 1990, 2000, and 2010. However,
numbers of natural passive parks did not differ significantly among census block
46
groups based on household income in 1990. Numbers of natural passive parks
increased significantly from 1990 to 2000 for census blocks that were below the
median for educational attainment and home value, but they did not increase
significantly from 2000 to 2010 for census block groups that were below the median
for educational attainment, home value, and household income (Table 7).
The area of natural passive parks was significantly higher (roughly twice) for
census block groups that were above the median in household income in 1990, 2000,
and 2010, for those that were above the median for education in 1990, and for those
that were above the median for home value in 1990 and 2000. However, the area of
natural passive parks did not differ significantly among census block groups based on
educational attainment from 2000 to 2010. There were no significant changes over
time in natural passive park area for census block groups that were above the median
for educational attainment. However, natural passive park area increased
significantly from 1990 to 2000 (but not from 2000 to 2010) for census block groups
that were above and below the median for home value and household income (Table
8).
Table 8 Mean number of natural passive parks within 800m of census blocks groups
above or below Seattle median for each socioeconomic category 1990 – 2010
47
Table 9 Group means for natural passive park area within 800m of census block
groups above or below Seattle median for each socioeconomic category 1990-2010.
4.6. Access to Multiuse Parks and change over time by education, home value, and
income
Overall, the numbers of multiuse parks was significantly higher for (2 parks
vs. 1 park) for census block groups that were above the city median for education,
home value and household income in 1990, 2000, and 2010. Numbers of multiuse
parks did not increase significantly from 1990 to 2010 for census block groups that
were above or below the median for educational attainment and household income;
however they did increase significantly from 2000 to 2010 for census block groups
that were below the median for home value (Table 9).
The area of multiuse parks was significantly higher (roughly twice) for census
block groups that were above the median for education, home value and income in
1990, 2000, and 2010. There were no significant changes in multiuse park area from
1990 to 2010 (Table 10).
48
Table 10 Mean number of multiuse parks within 800m of census blocks groups above
or below Seattle median for each socioeconomic category 1990 – 2010
Table 11 Group means for multiuse park area within 800m of census block groups
above or below Seattle median for each socioeconomic category 1990-2010.
49
4.7. Trends in park access over time by socioeconomic group
Differences in park access in all three decades were largest based on
household income (Tables 5 to 10). In 1990, census block groups that were above the
median for household income had access to 1.7 fewer recreation parks and 0.5 more
natural passive and multiuse parks compared to census block groups that were below
the median for household income. In 2010, census block groups that were above the
median for household income had access to 1.3 fewer recreation parks, and 0.7 more
natural passive parks, and 0.4 more multiuse parks compared to census block groups
that were below the median for household income (Figure 6).
However, park access differences in terms of area tell a different story. In
1990, census block groups that were above the median for household income had
access to more area of all types of parks compared to census block groups that were
below the median for household income (Figure 7). Moreover, by 2010, although
census block groups that were above the median for household income had access to
slightly less area of recreation parks, they had access to the same area of natural
passive parks, and to increased area of multiuse parks (Figure 7).
Figure 6. Differences in the average numbers of parks available to above- versus
below-median census block groups based on household income over time in Seattle.
50
Figure 7. Differences in the average area (m) of parks available to above- versus
below-median census block groups based on household income over time in Seattle.
4.8. Relationship of park access to race
Census block groups that were below the median for educational attainment,
home value, and income had significantly higher percentages of blacks (by 10
percentage points) compared to census block groups that were above the median, in
1990, 2000, and 2010 (Table 11). Census block groups that were below the median
also had significantly higher percentages of Asians (by 6 to more than 10 percentage
points) compared to census block groups that were above the city median in 1990,
2000, and 2010 (Table 12). From 1990 to 2000, the percentage of blacks and Asians
above and below the median for all socioeconomic characteristics did not
significantly change. However, from 2000 to 2010, the percentage of African
American means decline for above median for all characteristics declined, in contrast
Asians slightly increased in census block groups above the median for all
characteristics (Table 11 and 12).
51
Table 12 Group means for percentage of African American racial group belonging
each census block groups that is above or below the median for each socioeconomic
category 1990-2010.
Table 13 Group means for percentage of Asian racial group belonging each census
block groups that is above or below the median for each socioeconomic category
1990-2010.
52
5. Discussion
5.1. Park access and socioeconomic variables in Seattle
This study showed that park access varied significantly among socioeconomic and racial groups in Seattle, and in some cases differences in access were
enhanced, rather than reduced, over time. We expected to find that census block
groups that were below the median for educational attainment, home value, and
income had greater access to the number and area of recreation parks within 800m.
As expected, we found that access varies as a function of educational attainment,
home value, and household income in Seattle. Census block groups that were below
the median for educational attainment and home value had greater access to the
number of recreation parks within 800m in 2010, but not in 1990 and 2000, when
access was equal. Although census block groups that were below the median for
household income had access to a greater number of recreation parks within 800m,
they did not have access to more recreation park area. Instead, recreation park area
access did not differ between census block groups that were above and below the
median for household income.
expected above the city median census block groups for educational
attainment, home value, and household income would have greater access to natural
passive and multiuse parks. We did not find support for census block groups above
the city median for educational attainment having greater access to natural passive
parks. Instead we found that census block groups above and below the city median
for educational attainment had equal access to the number and area of natural passive
parks within 800m.
5.2. Methodological issues
As suggested by Boone et al (2009) and Zhou and Kim (2013), using park
acreage as a measure of access can tell a different story than solely measuring access
by counts of parks within a certain distance and can reflect the variability in park
access. In Baltimore, although Blacks had more parks within a quarter mile, they had
53
access to less area of parks than white neighborhoods (Boone et al 2009). In this
study suggest that there is no difference in terms of counts compared to area.
Previous authors have ignored park types, which can obscure important
differences. In a study of Milwaukee, Wisconsin, by distinguishing public from
private parks, Heynen et al (2006) were able to show that canopy coverage in
neighborhoods was based on purchasing ability, rather than lack of public funding for
trees. Examining park access by park types also reveals the different kinds of
benefits parks provide to different socioeconomic groups. For example, in this study
above-median census block groups had access to more natural passive and multiuse
parks, but fewer recreation parks.
5.3. Changes in park distribution over time - the role of park investment
Park investment in Seattle from 1990 to 2010 had the effect of maintaining
differences in park access based on socioeconomic variables. We expected to find an
overall increase in park area over time, and creation of new parks, or expansion of
existing parks, in areas formerly lacking parks. As expected, park area increased for
all park types from 1990 to 2010. However, no new recreation parks or multiuse
parks were added to areas formerly lacking those types of parks. In 1990, recreation
parks were clustered in the northwest, east, and southwest sections of the city, but
they were scarce in the southeast region. From 1990 to 2010, most recreation parks
added to the Seattle park system had at least one other recreation park within 800m.
In 1990, multiuse parks were clustered in the central and northern sections, but they
were scarce in the southern sections of the city. Over half of the multiuse park
parcels added to the park system from 1990 to 2000 and 70% of the parcels added
from 2000 to 2010 expanded pre-existing parks in the central and northern sections of
the city.
In contrast, new natural passive parks were added to areas formerly lacking
natural passive parks. Most of these additions were expansions of existing parks, but
in some cases, new park parcels were established in areas formerly lacking natural
passive parks. For example, the West Duwamish Greenbelt was extended from 15 to
120 acres between 1990 and 2000. The East Duwamish Greenbelt also was extended.
54
These additions shifted the overall distribution of natural passive parks toward areas
that formerly lacked parks of this type and is linked to Seattle’s prioritization due to it
being a critical link in the wildlife habitat corridor.
The lack of change in spatial patterns of recreation and multiuse parks,
compared to the expansion of natural passive parks, is attributable to the plans,
funding, and park policies of the City of Seattle and King County. In 1988, the city
of Seattle adopted a revised Comprehensive Plan and policies involving open space
(enacted in 1990), which established long-term open space goals. A Comprehensive
Open Space Program was developed, which included: (1) establishing a continual
funding source for acquisition of land, (2) seeking funding for maintenance expenses,
and (3) use of non-purchasing options such as donations, land trades, etc. The plan
also addressed public and private greenspace development, by targeting the
acquisition of areas accessible to a large number of people or having special features
such as streams, wooded hillsides, and/or historic landscapes (Resolution 27852
1988). The Open Space and Trails program targeted 27 greenbelts and natural areas,
and by 1998 about 600 acres of land had been preserved. The acquisition of land was
achieved through outright purchases and non-purchasing methods including donations
and land transfers, as well as funds from the 1989 King County Open Space and
Trails Bond, passed in 1989. This bond initially provided Seattle with $41 million
but the budget grew to nearly $100 million (Seattle Parks and Recreation 2010).
Thus, these policies of the city of Seattle and King County, which focused on growth
of natural passive park lands, explain the trends in natural park distribution observed
in this study.
Spatial and temporal patterns of access to multiuse parks also can be
explained in part by patterns of investment in parks. Census block groups above the
city median for all variables had greater access to multiuse parks in 1990, 2000, and
2010. Over 80% (113 of 139 acres) of new multiuse park parcels added to the 2,405
acres of multiuse parks in the Seattle park system from 1990 to 2010 extended
previously existing parks in areas mostly bordering the Puget Sound and Lake
Washington coastlines, whose neighboring census block groups are above the median
for all socioeconomic characteristics. Also, 80% (40 of 58 acres) of multiuse
55
parkland added to the park system in 2000 to 2010 extended previously existing parks
located along the shoreline that serve wealthy, highly educated communities. This
suggests that investment in multiuse parks might be disproportionately beneficial to
communities that are wealthy and highly educated. This topic is explored further in
Chapter 4.
5.4. Disparities in park access and environmental justice
This study found that census block groups with below-median levels of
household income and home value consistently had less access to multiuse natural
and natural passive parks, and greater access to recreation parks than above-median
census block groups, and that these patterns did not change from 1990 to 2010.
Because census block groups that were below the median for educational attainment,
home value and household income had significantly greater proportions of minorities
than census block groups that were above the city median, census block groups with
high proportions of minorities have significantly less access to multiuse and natural
passive parks in Seattle.
Many studies have shown that ‘green public open spaces’ (similar to natural
passive parks) are smaller in poor areas (Talen 1997; Cutts et al. 2009). This study
provides a novel finding that although less area of natural passive parks is available to
census block groups that were below the median for home value and income,
nevertheless access to natural passive park acreage increased significantly from 1990
to 2010 from 9 to 21 acres, 7 to 18 acres, and 5 to 13 acres, in census block groups
that were below the median for education, home value and income, respectively.
The finding that greater numbers of recreation parks tend to be nearer to
communities below the median could be considered an injustice. Based on the
typology natural passive and multiuse parks provide more green amenities than do
recreation parks. For example if we understand certain park types to provide more
ecosystem services such as air and nose filtering based on the percentage of canopy
coverage, many recreation parks would not meet the criteria for providing these
services.
56
5.5. Changes in park access and implications for gentrification
Surprisingly, both above and below city median census block groups for
educational attainment had equal access to the number and area of natural passive
parks within 800m. The finding that access to natural passive parks varied by income
and home value, but not by education level, may indicate that educational status and
park access are indicators of ongoing demographic changes in neighborhoods. In this
study, natural passive parks expanded in south Seattle, where many census block
groups have below-median home value and income, but where education levels may
be rising. This is similar to a study which found that census block groups with higher
education and higher income were typically associated with a greater tree canopy (as
expected to see in natural passive park types) in six cities in Illinois (Zhou and Kim
2013).
Park distribution is a legacy of past demographics combined with recent park
development and investment, indicative of a dynamic city system. Differing park
types appear to be part of this process and are connected to income, home value and
race. Although adding multiuse or natural passive parks in areas previously lacking
access might benefit the neighborhood, it may also be a problem if it leads to
increases in home value and income that displace residents and change the
fundamental characteristics of the neighborhood. This process is called gentrification
and is explored further in Chapter 3.
5.6. Implications for park planning
Despite an overall increase in access to parks in Seattle, and a slight reduction
in the disparity of access to natural passive parks, the overall disparity of access to
parks between above- and below-median groups, and between white vs. minoritydominated neighborhoods, did not change from 1990 to 2010. Most additions to park
area were quite small, especially for multiuse parks, which partly explains why park
access has not changed much over time. However, constraints on available land also
may explain these results. Seattle’s land area may not allow for large parks to be
established in areas that were previously lacking larger parks. Limited land area in
highly developed areas may preclude the establishment of all parks except small
57
recreation or multiuse parks. Park planning focused on park additions and
improvements in areas that are lacking park access could address some of the
inequities.
All in all it is important to consider the equity of distribution in the planning
process for a number of reasons. The disparity in park access raises questions about
environmental justice. Traditional environmental justice literature has focused on the
distribution of environmental bads, however the literature has developed to also
include fair distribution of environmental goods such as quality air, water, and parks.
The distribution of environmental goods may also be linked to issues surrounding
gentrification and environmental gentrification. The results of this analysis appear to
show a trend that educational attainment and home value increased in areas where
income was already above the mean. This may suggest that economic factors like
income may lead to improvements in education and home value which have been
suggested to be related with park access. Access to multiuse and natural passive
parks was higher for the white racial group, which also had higher home value,
educational attainment, and household income, raising the question of whether
investments in park improvements led or followed the changes in educational
attainment, home value, and income. The environmental gentrification theory should
be tested to further understand how park access may or may not impact communities.
6. Conclusion
The results of this study clearly show that access to different park types is not
uniform across the city of Seattle. The socioeconomic fabric of certain census block
groups determines access to different park types. Access to parks within walking
distance is not evenly distributed throughout the city. On average greater numbers of
recreation parks within walking distance are available to neighborhoods having lower
incomes, lower home values and lower percentages of college degree holders. In
contrast, natural passive and multiuse parks are more available to neighborhoods with
socioeconomic characteristics above the city median. These results are similar to
other studies, which found disparities in access to parks between neighborhoods of
varying socioeconomic status (Gobster 2002, Timperio et al 2007, and Boone et al
58
2009). This and other studies have also confirmed that park access differs among
racial groups. Blacks and Asians more frequently inhabit census block groups that
are below the city median for educational attainment, home value, and household
income, have greater access to recreation parks, but lack access to multiuse and
natural passive parks.
Overall, the distribution of parks and varying access to each park type in
Seattle brings light to the complex interactions between socioeconomic status and
planning efforts for the placement of newly added parks over time. It can be
challenging to change or reduce disparities in park access over time due to limits of
available land, zoning, and other budgetary issues.
59
7. References
Banzhaf, H. and R. Walsh. 2004. Testing for environmental gentrification: Migratory
Responses to changes in environmental quality. Paper for AERE Workshop. Estes
Park, CO
Bolund, P. and Hunhammar, S. 1999. Ecosystem services in urban areas. Ecological
Economics 29: 293-301.
Burgess, J., C.M. Harrison, M. Limb. 1988. People, Parks and the Urban Green: A
Study of Popular Meanings and Values for Open Spaces in the City. Urban Studies.
25, 455-473.
Chiesura, A. 2004. The role of urban parks for the sustainable city. Landscape and
Urban Planning. 68 (1): 129-138.
Conway, H. 2000. Parks and people: the social functions. In: Wousstra, J.,
Fieldhouse, K. (Eds.), The Regeneration of Public Parks.
Cranz, G., and Boland, M. 2004. Defining the sustainable park: A fifth model for
urban parks. Landscape Journal. 23: 2-4.
Cutts, B. Darby, B. Boone, C. G., Brewis, A. 2009. City Structure, obesity, and
environmental justice: an integrated analysis of physical and social barriers to
walkable streets and park access. Social Science & Medicine 69: 1314-1322.
Crompton, J.L. 2001. The Impact of Parks on Property Values: A Review of the
Empirical Evidence. Journal of Leisure Research. 33; 1-31.
Ernst and Young 2003 cited in Voicu, I., and V. Been. 2008. The effect of community
gardens on neighborhing property values. Real Estate Economics. 36(2): 241-83.
Espey, D.J. and Owusu-Edusei, K.J. 2001. Neighborhood Parks and Residential
Property Values in Greenville, South Carolina. Journal of Agricultural and Applied
Economics. 33: 487-492.
Federal Highway Administration. 2016. Recreational Trails Program.
http://www.fhwa.dot.gov/environment/recreational_trails/ last accessed 17 February
2016/
Forsyth, A. 2000. Analyzing public space at a metropolitan scale: Notes on the
potential for using GIS. Urban Geography 21: 121-47.
Gobster, H. 2002. Managing Urban Parks for a Racially and Ethnically Diverse
Clientele. Leisure Sciences: An Interdisciplinary Journal 24 (2): 143-159.
60
Hartig, T., M. Mang, G. Evans. 1991. Restorative effects of natural environments
experiences. Environmental Behavior. 23: 3-26.
Harnik, P. 2004. Parks: How far is too far? Planning. 9-11.
Heynen N., Perkins, H.A., Roy, P. 2006. The political ecology of uneven urban green
space: the impact of political economy on race and ethnicity in producing
environmental inequality in Milwaukee. Urban Affairs Review 42: 3-25.
Heynen, N. and G. Lindsey. 2003. Correlates of urban forest canopy cover.
Implications for local public works. Public Works Management & Policy. 8 (1): 3347.
Kaplan, R. 1983. The analysis of perception via preference: a strategy for studying
how the environment is experienced. Landscape Urban Planning. 12: 161-76.
King County GIS Center. 2014. GIS Data Portal.
http://www5.kingcounty.gov/gisdataportal/ (last accessed 07 June 2014).
Kornblum, W. 1978. The Psychology of Space in Taylor, L. ed. Urban Open
Spaces.London: Academy: 15-16.
Lindsey, G., M. Maraj, and S. Kuan. 2001. Access, equity, and urban greenways: An
exploratory investigation. Professional Geographer 53:332–46.
Liu, S. X., and X. A. Zhu. 2004. Accessibility Analyst: An integrated GIS tool for
accessibility analysis in urban transportation planning. Environment and Planning BPlanning & Design 31:105–24.
Lutzenhiser, M. and Netusil, N.R. 2001. The Effect of Open Spaces on a Home’s Sale
Price. Contemporary Economic Policy. 19: 291-298.
Marcus, C.C and Francis, C. 1998. People Places Design guidelines for urban open
space. NewYork, NY: John Wiley & Sons, Inc.
Maas, J., R. A. Verheij, S. de Vries, P. Spreeuwenberg, F. G. Schellevis, and P. P.
Groenewegen. 2009. Morbidity is related to a green living environment. Journal of
Epidemiology and Community Health. 63(12): 967–73.
McGee, H. 2008. Seattle’s Central District, 1990—2006: Integration or
Displacement?
The Urban Lawyer 39:167
Moore, M. and Woodrow, J. 1982. Recreation and Urban Park Planning; Problems
and Prospects. Journal of Environmental Systems. 11 (4): 315-23.
61
Nicholls, S. 2001. Measuring the accessibility and equity of public parks: A case
study using GIS. Managing Leisure 6: 201-19.
Talen, E. 1997 The social equity of urban service distribution: An exploration of park
access in Pueblo, Colorado, and Macon, Georgia. Urban Geography 18 (6): 521-41.
____ 2010. The spatial logic of parks. Journal of Urban Design 15 (4): 473-491.
Seattle Parks and Recreation. 2010.Parks and open space acquisition report.
Memorandum.
http://www.seattle.gov/parks/parkboard/briefings/park_levy_acquisition_program.pdf
. Last accessed 16 February 2016.
Swyngedouw, E. 1999. Modernity and hybridity: Nature, regeneracionismo, and the
production of waterscape, 1890 – 1930. Annals of the Association of American
Geographers 89 (3): 443-465.
Resolution 27852: Comprehensive Open Space Program. City of Seattle Urban
Development. Office of the City Clerk.
http://clerk.ci.seattle.wa.us/~scripts/nphbrs.exe?s1=&s3=27852&s2=&s4=&Sect4=A
ND&l=20&Sect2=THESON&Sect3=PLURON&Sect5=RESNY&Sect6=HITOFF&d
=RESF&p=1&u=/~public/resny.htm&r=1&f=G Last accessed 16 February 2016.
Troy, A. and Grove, J.M., 2008. Property Values, Parks and Crime: A Hedonic
Analysis in Baltimore, MD. Landscape and Urban Planning. 87: 233-45.
Ulrich, R.S. 1981. Natural versus urban sciences: some pyscho-physiological effects.
Environment Behavior. 13: 523-56.
Voicu, L. and Been, V. 2008. The Effect of Community Gardens on Neighboring
Property Values. Real Estate Economics. 36, 2: 241-283.
Washington State Geospatial Data Archive. 2007. http://wagda.lib.washington.edu/
(last accessed 10 June 2014).
Zhang, X., Lu, D., Holt, J.B. 2011. Modeling spatial accessibility to parks: a national
study. International Journal of Health Geographies. 10 (31): 1-14.
Zhou and Kim 2013. Social disparities in tree canopy and park accessibility: A case
study of six cities in Illinois using GIS and remote sensing. Urban Forestry & Urban
Greening.12: 88-97.
62
8. Appendix
The frequency and distribution of educational attainment for census block groups in
Seattle Washington for the 1990, 2000, and 2010 census years.
63
The frequency and distribution of median household income for census block groups
in Seattle Washington for the 1990, 2000, and 2010 census years.
64
The frequency and distribution of median home value for census block groups in
Seattle Washington for the 1990, 2000, and 2010 census years.
65
Chapter 3
Spatial patterns of gentrification in Seattle: multifaceted, dynamic interactions
Abstract
Mapping of gentrification is a necessary first step in order to test hypotheses about
causes and consequences of gentrification. The availability of digital spatial data,
including census data as well as spatially explicit media coverage, provides the
opportunity to develop and test an approach to mapping gentrification in Seattle over
the period 1990 to 2010. Gentrification was defined as change from below to above
the median value for education, income, and home value in a census block group over
the period 1990 to 2000 or 2000 to 2010. These patterns were compared to changes
in proportions of minorities and perceptions of gentrification in the media over those
time periods. Nine percent of census block groups met the definition of gentrification
in each of the two decades. Gentrification was clustered in central, north and west
Seattle. In Central Seattle these changes were accompanied by large declines in the
proportion of African Americans in the population. However, gentrification in terms
of education, income, and home value also occurred in areas that did not have large
proportions of minorities. Gentrification was spatially contagious: it was more likely
to occur in census block groups that were adjacent to previously gentrified census
block groups. Media perceptions of gentrification were reported in places that met
the definition of gentrification, but also in areas experiencing changes in minorities,
but which did not yet meet the definition of change in terms of education, home
value, and income. These findings demonstrate that gentrification is a multidimensional, spatially contagious, dynamic process, and gentrification may not
always lead to displacement of minorities. More work is needed to understand the
causes of gentrification, especially the role of environmental amenities.
Keywords: gentrification,
66
1. Introduction and Research Questions
Urban areas have strongly defined spatial patterns of wealth and race, which are
often correlated, and they are dynamic. While some parts of an urban landscape may
be invariant over time, other parts may experience rapid and dramatic change.
Neighborhoods experience change over time in demographic characteristics, such as
income, home value, and education levels, as well as racial and ethnic composition.
When a relatively poor neighborhood experiences increasing wealth, and this forces
minorities to move out, the process is described as "gentrification" (Glass 1964,
Smith 1982; Kendig 1984; Williams 1984). Gentrification is considered to be a form
of social injustice, because it disproportionately disadvantages poor and minority
groups. It may also lead to feelings of loss, anger, and fear in neighborhoods that are
gentrifying (Smith 1996). Moreover, urban planning significantly influences spatial
patterns of gentrification (Smith 2002). Therefore, explicit spatial information about
the patterns of gentrification is useful both for understanding its causes and
consequences and for urban planning.
Despite widespread discussion of gentrification in the scholarly literature and
in the popular media, few standard approaches exist to map and describe
gentrification. Mapping of gentrification is a necessary first step in order to test
hypotheses about causes and consequences of gentrification. Maps of gentrification
have been constructed based on quantitative vs. qualitative information and objective
vs. subjective classification. For example, in Seattle, gentrification has been mapped
based on changes in census data including population, race, persons with college
degrees, income, age, housing values, and differences in housing types and
composition (Morrill 2003). Other studies map gentrification based on qualitative
information such as cultural changes and shifts in demographics, such as ethnic
culture change over to white culture (Anderson 1990; Maurrasse 2006)
Seattle, in the state of Washington, USA, has experienced demographic
change in many neighborhoods, and some of this change has been described as
gentrification. Seattle’s Central District transformed from predominantly working
class African American neighborhoods to mostly professional White, Asian
American, and African Americans from 1990 to 2000 (McGee 2008). More recently,
67
Seattle was described as one of the “most gentrified cities in America” (Business
Insider.com). In February 2014, The Seattle Times reported that “two masked men
blocked a Microsoft employee shuttle on Capitol Hill…condemning the
neighborhood’s transformation into an ‘upscale yuppie playground’…[with]
skyrocketing rents that are displacing less-affluent, longtime residents” (The Seattle
Times).
The availability of digital spatial data, including census data as well as
spatially explicit media coverage, provides the opportunity to develop and test an
approach to mapping gentrification in Seattle. This study has the following
objectives:
(1) Develop a definition of gentrification based on socioeconomic data (income,
home value, education, race) and apply this definition to map gentrification in
Seattle from 1990 to 2000, and 2000 to 2010.
(2) Test how patterns of change in income, home value, and education are related to
spatial patterns of change in African American and Asian populations and to
media reports of gentrification.
(3) Infer how results help refine understanding of gentrification and its relationship to
patterns of socio-economic and demographic change in Seattle.
Hypothesis 1: Gentrification is more likely in neighborhoods with relatively low
income, home value, and education, as well as high proportions of minorities.
Socioeconomic characteristics such as low income, home value and educational
attainment have been linked to gentrification (Vigdor 2002).
Hypothesis 2: Gentrification is more likely in locations adjacent to areas that
gentrified in a previous time period.
Based on Tobler’s first law, it is likely that gentrification would occur in areas
adjacent to where it previously occurred.
Hypothesis 3: Increases in income, home value, and education coincide with declines
in proportions of minorities.
68
Hypothesis 4: Media accounts perceive gentrification in places that experience
declines in populations of minorities, but not in places that only experience
changes in education, home value, and income.
Qualitative accounts may be equally, if not more effective than quantitative analysis
of socioeconomic variables for identifying gentrification (Barton 2016).
2. Study Site
Seattle, Washington, is the largest city in the Pacific Northwest with an area of
217 sq km and a population of 608,255 (US Census, 2010). The racial fabric of the
city is mostly Whites (69.5 percent), Asians (13.8 percent), and Blacks (7.9 percent).
American Indian and Alaska Native, Native Hawaiian and other Pacific Islanders, and
two or more races together make up the remaining 8.8 percent. In contrast to the state
of Washington as a whole, Seattle is highly educated. As of 2010, 57.4 percent of the
population age 25 and older held a bachelor’s degree or higher, compared to only
31.9 percent in the state of Washington and 28.8 percent in the US (United States
Census Bureau, 2009-2013). However, less than half of the population of Seattle
(46.8 percent) owns a home, compared to 63.2% in the state of Washington and 64.9
percent in the US. The proportion of households below the poverty line is 13.6
percent, which is comparable to the rest of the state of Washington.
3. Data sources and Methods
This study involved (1) developing a definition of gentrification based on
socioeconomic data (income, home value, education, race) and applying this
definition to map gentrification in Seattle from 1990 to 2000, and 2000 to 2010; (2)
comparing this map to spatial patterns of change in African American populations;
and (3) conducting a content analysis to identify the spatial coordinates of media
reports of gentrification.
3.1 Data sources
Racial and socioeconomic data were obtained from the 1990, 2000, and 2010
US Census. The data were acquired from the National Historical Geographic
69
Information System (NHGIS) through the Minnesota population center, University of
Minnesota.
The geographic unit of analysis for all census related data was the census
block group. A block group is a statistical subset of a census tract, and generally
contains between 600 and 3,000 people (US Census Bureau 2012). Each block group
is uniquely identified allowing it to be represented geographically with a GIS. As of
2010, Seattle had a total of 481 block groups (Figure 1). In 2000 there were 570
census block groups, and in 1990 there were 630. Due to the changing boundaries of
census block groups over time, this study aggregated the 1990 and 2000 census block
group data to the 2010 census block group area based on an area weighting
methodology.
Three socioeconomic variables were selected for this study, based on previous
research in the quality of life, environmental justice, environmental gentrification and
environmental justice literature (e.g. Jensen 2008, Sieg et al., 2004 and Boone 2009).
Median household income and median home value were used to represent the
economic status of each census block group. The percentage of residents 25 and
older with a bachelor's degree or higher was used to represent the education status of
each census block group (Table 1). Racial groups were characterized based on the
proportion of each census block group that was African American or Asian.
The socioeconomic census data in .csv files and census block groups in
shapefile format were downloaded from the NHGIS website. The .csv files and
shapefiles were imported into the ArcMap data frame using ArcGIS10.2.2. The .csv
files and the shapefiles were joined together using the join function to attribute the
socioeconomic and racial data for each census block group based on its unique
identifying object ID number.
70
Table 1 Socioeconomic variables collected by the US Census Bureau (Census
Bureau, 1990, 2000, 2010). These variables were used to analyze the process of
socioeconomic change from 1990 – 2010.
3.2. Spatial data analysis
The overall median household income, median home value, and educational
attainment were calculated for the City of Seattle for 1990, 2000, and 2010 (Table 2).
Each census block group was classified as above (labeled as A) or below (labeled as
B) the overall city median for each of these variables in each time period. For each
socioeconomic variable (education, home value, household income), the trajectory
relative to above- or below-median status of each census block group was determined
from 1990 to 2000 and from 2000 to 2010. There are eight possible trajectories of
transition from 1990 to 2010 (Table 3).
71
Table 2 City of Seattle overall median for selected socioeconomic characteristics for
each decennial census year (1990, 2000, and 2010).
Table 3 The transition trajectory for census block groups above or below the city
median in for each census year.
3.3 Mapping gentrified census block groups
Seven trajectories of change were possible for a census block group in each time
period (1990 to 2000, 2000 to 2010) based on three socioeconomic variables;
education, home value, and income (Figure 1). A gentrified census block group was
defined as having at least two socioeconomic characteristics that transitioned from
below the median in t1 to above the median in t2. Six other trajectories were not
72
identified as gentrification: (1) no change, (2) an increase in one variable
compensated by a decrease in another variable, (3) decrease from above to below
median in two or more variables, (4) change from above median in three variables to
above median in only two variables, (5) change from above-median in two variables
to above-median in three variables, (6) change from below-median in three variables
to below-median in two variables. These trajectories were mapped for the two
periods, 1990-2000 and 2000-2010.
Figure 1. Example transition matrix showing the numbers of census block groups
that experienced seven possible trajectories of change from 1990 (t1) to 2000 (t2) in
Seattle, Washington. 3-Below (3-above) means that a census block group was below
(above) the median in education, home value, and household income in that time
period. 2-Above or 2-Below means that two variables were above (below) the
median. Hmval = home value, Edu = educational attainment, and Inc = household
income. For example, a census block group classified as 2-Above (Hmval and Edu)
had median home value and educational attainment that were above the median for
the City of Seattle in that period, but the median household income was below the
median for all households in the City of Seattle.
73
3.4 Change in African American populations
The change in percentage of African Americans was calculated (percent in time 1
minus percent in time 2) for each census block group for the time periods 1990 to
2000 and 2000 to 2010. The resulting layer was overlaid on the map of gentrification
for the corresponding time period.
3.5. Content analysis on select neighborhoods of gentrified census block groups
This study used a semi-structured media content analysis to identify locations of
perceived gentrification from news sources and blog websites from 1990 – 2010.
Content analysis is an important technique used in social sciences and communication
research to infer meanings and identify characteristics within the content (Holsti
1969). The content analysis in this study was conducted to determine how and where
gentrification is perceived by the media and the community and to validate the
quantitative spatial analysis of census data.
Data for the content analysis were acquired through a search of the LexisNexis
Academic and Google databases, available via Oregon State University Libraries and
Google. The databases were searched using names of Seattle neighborhoods (ie:
Central District), as well as terms related to gentrification, displacement, and
affordable housing. The following search terms were used for the initial search:
gentrification AND Seattle, Washington OR affordable housing, AND "Central
District", OR "Ballard", OR "Capitol Hill" OR "West Seattle." All identified articles
were downloaded and combined into a single word document. This generated a
sample of 118 articles. Next, the articles were filtered to identify all that were related
to Seattle and events related to gentrification; this reduced the sample to
approximately 100 articles. Utilizing word cloud web applications and general word
and phrase search methods, the content was analyzed to identify emerging patterns
and themes. Instances of perceived gentrification in the articles included (1) changes
in neighborhoods due to gentrification, (2) neighborhood changes based on shifting
demographic, (3) other anecdotal accounts that signify social, economic, and cultural
changes within and around neighborhoods. The spatial coordinates of perceived
gentrification were identified by place names and street addresses mentioned in the
74
articles. Using Google Earth, I searched for the place names or addresses and created
pin points for each place. Google saves place pin points in .kml format, which I then
converted into a shapefile in ArcGIS 10.2.2.
4. Results
4.1 Changes in spatial patterns of education, home value, and household income
For educational attainment, in 1990 more than half of the census block groups
were below the median, but in 2000 the census block groups were almost evenly split,
and by 2010, more than half of the census block groups were above the median for
educational attainment (Table 4). For home value consecutively each census year
more than half of the census block groups were below the median (Table 4). For
median household income, in 1990, 2000, and 2010 more than half of the census
block groups were above the city median (Table 4).
Table 4 The percentage of census block groups categorized as greater than or less
than the city of Seattle median for each socioeconomic characteristic by year.
In 1990 census block groups above the city median for household income
were concentrated in north Seattle, and along the Puget Sound and Lake Washington
(Figure 2-A). Most of the census block groups below the city median for household
income were found in south Seattle. This trend continued into 2000 and 2010 (Figure
2-B and C). Forty-two percent of census block groups were above the median in
1990, 2000, and 2010 (AAA); these were located along the Puget Sound and Lake
Washington. Census block groups that were below the median in all three time
75
periods (BBB) made up 29 percent of the census block groups and were clustered in
South Seattle and in North Seattle (Table 5-A and Figure 5-A).
In 1990, census block groups that were above the city median for home value
were mostly found along the Puget Sound and Lake Washington areas. This pattern
was fairly consistent for each census year observed (Figure 3). Census block groups
that were above the median in all three time periods (AAA) made up 28.9 percent of
the census block groups. They were found in large clusters along the Puget Sound
and Lake Washington areas as well as inland in North Seattle near Green Lake and
Lake Union (Figure 5-B). Census block groups categorized as BBB made up 43.2
percent of the census block groups (Table 5-B). Large clusters of census block
groups below the median in all three time periods (BBB) were found in both north
and south Seattle (Figure 5-B).
Educational attainment had a steady increase in the percentage of census
block groups that were above the city median from 1990 to 2010 (Table 4).
However, this increase is mostly found in the northern sections of the city whereas
most of south Seattle remained below the city median from 1990 to 2010 (Figure 4).
Thirty-six percent of census block groups were above the median in all three time
periods (AAA), and these were located in north Seattle specifically toward the east
side of the city with some clustering along the north end of the city along the Puget
Sound. Census block groups categorized as BBB made up 37 percent of the census
block groups and were found in in large clusters in South Seattle, and one large
cluster in North West Seattle. There also appears to be a contagion effect of clusters
of census block groups categorized as BBA and BAA that are adjacent to BBB census
block groups in both North and south Seattle (Figure 5-C).
76
Table 5 The percentage of census block groups belonging to each transition for each
socioeconomic characteristic from 1990 to 2010.
77
Figure 2 The spatial distribution of census block groups categorized as being greater
than or less than or equal to the city of Seattle median for median household income
from 1990 – 2010.
78
Figure 3 The spatial distribution of census block groups categorized as being greater
than or less than or equal to the city of Seattle median for median home value from
1990 – 2010.
79
Figure 4 The spatial distribution of census block groups categorized as being greater
than or less than or equal to the city of Seattle median for educational attainment from
1990 – 2010.
80
Figure 5.The graphical representation of the transition matrix showing the locations
of each census block group above or below the city median for each socioeconomic
category from 1990 to 2010. .
81
4.2 Spatial patterns of socioeconomic change and gentrification, 1990 to 2000
The analysis of trajectories of change in three socioeconomic variables
(education, home value, and income) reveals complicated spatial patterns in Seattle
from 1990 to 2000 (Figure 6) and 2000 to 2010 (Figure 7).
From 1990 to 2000, 9% of all the census block groups gentrified, based on an
increase from below to above-median in two or more socioeconomic variables:
education, home value, and household income. This pattern appears to be clustered,
and most census block groups that gentrified during this time period appeared to be
adjacent to at least two other gentrified census block groups. The clustered pattern is
most apparent in Northwest Seattle, among the Ballard and Freemont neighborhoods.
Other clusters of gentrified census block groups were located in the Capitol Hill,
Downtown, and West Seattle neighborhoods (Figure 6).
Overall, 61.3% of census block groups had no change and were evenly dispersed
throughout the city. Census block groups that were below the median in all three
categories in 1990 and remained below the median in all three categories in 2000
were located in both North and South Seattle, with the majority of occurrences in
South Seattle (areas labeled BBB in Figure 6). Census block groups that were above
the median in all three categories from 1990 to 2000 were concentrated in North
Seattle along the Puget Sound and Lake Washington, and inland in Magnolia, Queen
Anne, and Capitol Hill neighborhoods (areas labeled AAA in Figure 6).
Census block groups that increased in one category and were compensated by a
decrease in another category made up 2.5% of all census block groups. They were
found mostly in North Seattle and bordered (on at least two sides) a census block
group that experienced no change. Six percent of the census block groups improved
from above median in two categories to above median in all three categories and
these occurred mostly in north and West Seattle. Eight percent of the census block
groups had an improvement from below median in three categories to below median
in only two categories. Three percent of the census block groups were above median
in all three categories and transitioned to above median in only two of three
categories. Lastly, 10% of census block groups showed deterioration from above
median to below median in two or more categories. Clusters of deteriorating census
82
block groups were found in north Seattle, downtown and Southeast and southwest
Seattle (Figure 6).
From 1990 to 2000, fifteen census block groups that gentrified had above-median
income in 1990. Eleven census block groups remained above the median for income
in 1990 and 2000 and transitioned to above the median in educational attainment by
2000. Most of these census block groups were located in North Seattle, but three
were in West Seattle, and one was in the Rainier Valley, Mount Baker neighborhood.
Three of the fifteen census block groups that led the socioeconomic transition with
above-median income transitioned to above-median home value in 2000. These
census block groups were located in North Seattle in the Freemont neighborhood,
Central Area, and West Seattle’s North Admiral neighborhood. The remaining
census block group that led with high income in 1990 transitioned to above-median
home value and educational attainment but income fell below the median in 2000.
This census block group was located in North Seattle in the Sand Point neighborhood.
From 1990 to 2000 fifteen census block groups that gentrified had educational
attainment above the median in 1990. Ten of these census block groups remained
above the median for educational attainment and transitioned to above-median home
value in 2000 and most were located in North Seattle’s Freemont, Wallingford and
the University District neighborhoods, while others were located in the Capitol Hill
neighborhood. Two of the fourteen census block groups remained above the median
for educational attainment and transitioned to above-median income in 2000; these
census block groups were located in the Phinney Ridge and University District
neighborhoods. Another two census block groups transitioned to below-median in
educational attainment and above-median in home value and income from 1990 to
2000; these census block groups were located in Downtown Seattle. The remaining
census block group located in Northeast Seattle in the Roosevelt neighborhood
transitioned all three categories above the median.
Only two of all gentrified census block groups was above-median home value in
1990. In 2000 both census block groups remained above the median for home value
and transitioned to above-median educational attainment. The census block groups
were in Ballard and Downtown neighborhoods.
83
Figure 6 The spatial distribution of socioeconomic change and gentrification in
Seattle from 1990 – 2000. (n=481)
84
4.3 Spatial patterns of socioeconomic change and gentrification, 2000 to 2010
From 2000 to 2010, 9% of all census block groups gentrified. Gentrification was
concentrated mostly in North Seattle in the Lake City, Northgate, Bitter Lake,
Broadview and Crown hill neighborhoods. Other clusters of gentrified census block
groups were apparent in the Central Area, Lower Queen Anne, Lake union and
Belltown neighborhoods (Figure 7).
From 2000 – 2010, 53.2% of census block groups experienced no change. Census
block groups that were below the median for all three socioeconomic variables in
2000 and 2010 were concentrated in southeast Seattle within the Beacon Hill
neighborhood (BBB in Figure 7). Census block groups that were above the median
for all three socioeconomic variables in 2000 and 2010 were concentrated along the
Lake Washington and Puget Sound coastlines and inland in Capitol Hill, Queen
Anne, Magnolia, Wallingford, Green Lake, Ravenna, and Bryant neighborhoods
(AAA in Figure 7).
Census block groups that increased in one category and were compensated by a
decrease in another category made up 4.2% of the census block groups. Census block
groups that improved from above the median in two categories to above the median in
all three categories made up 4.8% of the total and were found mostly in North Seattle
with some clustering in West Seattle. Census block groups that improved from
below-median in three categories to below median in two categories made up 7.7% of
the total and were evenly dispersed in throughout North and South Seattle. Census
block groups that experienced a small decline in socioeconomic status mostly
occurred in North and West Seattle. Census block groups that declined in
socioeconomic status from 2000 to 2010 made up 15.4% of the total and were located
in both North and South Seattle (Figure 7).
85
Figure 7 The spatial distribution of socioeconomic change and gentrification from
2000 to 2010.
86
From 2000 to 2010, thirteen census block groups that gentrified were above the
median for household income in 2000. Of these, nine census block groups remained
above the median for income and transitioned to above-median educational
attainment by 2010. Six of these census block groups were located in North Seattle’s
Ballard, Bitter Lake, Northgate and Lake City neighborhoods, and three were located
in West Seattle’s Fairmount Park and Gatewood neighborhoods. Three census block
groups remained above the median for household income and transitioned to abovemedian for home value by 2010; these were located in North Seattle, Central Area,
and the Industrial District neighborhoods. One census block group transitioned to
above-median home value and education, but to below-median income in 2010; this
census block group was located in West Seattle’s North Admiral neighborhood.
Two census block groups that gentrified from 2000 to 2010 had above-median
home value in 2000. These census block groups remained above the median for
home value and transitioned to above-median educational attainment in 2010; they
were located in Capitol Hill and West Seattle.
Eight census block groups that gentrified from 2000 to 2010 had above-median
educational attainment in 2000. Five of these transitioned to above-median income
by 2010. They were located in North Seattle’s Greenwood, Wedgewood and Queen
Anne neighborhoods. The remaining three census block groups transitioned to
above-median home value by 2010; they were located in the Queen Anne, Central
Area, and Fremont neighborhoods.
4.4 Racial changes in gentrified census block groups, 1990 – 2000
From 1990 to 2000, gentrified census block groups with high proportions of
African American or Asian populations were negatively affected by the changes
occurring the their neighborhoods. In 1990, high proportions of African Americans
were found in Central Area and Rainier Valley neighborhoods. In 1990 census block
groups G53003300088001 and G53003300089002 African Americans made up 74.9
and 56.1% of the population, but the percentages dropped to 44.1 and 42.7% by 2000,
coinciding with gentrification of these census block groups (Figure 8). In the Rainier
Valley, census block group G53003300101005 also gentrified, and the African
87
American population declined from 42.6% to 27.6% of the population from 1990 to
2000 (Figure 9). The Asian population within this census block group also decreased
from 18.4% in 1990 to 16.4% in 2000. Census block groups with low percentages of
Asian populations that gentrified did not experience significant changes in Asian
populations; most had 1-2 percentage point increases in Asian populations.
Figure 8 Change in the African American racial composition in the Central Area
from 1990 to 2000.
88
Figure 9 Change in the African American racial composition in the Rainier Valley
from 1990 to 2000.
From 1990 to 2000, census block group # G53003300053022, located in the
University District, gentrified. Over the same period Asians declined from 43.5% to
only 7.7% of the total population, African Americans decreased from 5.1% to 2.5%,
and Whites increased from 42 to 79.8% of the population in this census block group
(Figure 10). In census block group G53003300044004, which gentrified from 1990
to 2000, Asians increased from 13.3 to 18.3% of the population while African
Americans decreased from 3.5 to 2.0% (Figure 6).
89
Figure 10 Change in racial composition of the Asian racial group in the
University District from 1990 to 2000.
4.5 Racial changes in gentrified census block groups, 2000-2010
The most drastic declines in African American population occurred in five
gentrified census block groups located in the Central Area neighborhood in Seattle.
From 2000 to 2010 African Americans decreased from 45% to only 5% of the
population of census block group # G53003300079001 and from 26.7% to only 3% of
the population of census block group #G53003300077004 (Figure 11). The
remaining census block groups that gentrified in the Central Area neighborhood also
had decreases in the African American population, but the declines were smaller
(declines of -5.8, -6.1 , and -8.5 percentage points for census block group #’s
90
G53003300087002, G53003300077001, and G53003300087001, respectively)
(Figure 11).
Two of the five census block groups (census block groups G53003300079001
and G5330003300077004) in the Central Area experienced the greatest declines in
African American populations in the city over this period. Census block group
G53003300079001 was below the median for all three socioeconomic characteristics
in 2000, but by 2010 it transitioned to above the median for educational attainment
and home value. Census block group G5330003300077004 was above the median
for educational attainment in 2000, and it transitioned to above the median for home
value and income, while educational attainment fell below the median by 2010.
Figure 11 Change in the African American racial composition in gentrified census
block groups in the Central Area from 2000 to 2010.
91
Seven census block groups that gentrified from 2000-2010 had positive
changes in their African American populations. From 2000 to 2010 African
Americans increased from 4% to 16% of the population of gentrified census block
group #G53003300094004 in the Beacon Hill neighborhood (Figure 12-D). Four
census block groups that gentrified in West Seattle also experienced an increase in the
African American population. From 2000 to 2010, the African American population
increased from 4.1% to 9.0% in census block group # G53003300105002; from 2.2%
to 7.7% in census block group # G53003300096002; and from 2.2% to 5.7% in
census block group # G53003300097023 (Figure 12-C). Two census block groups in
North Seattle had positive changes in African American populations from 2000 to
2010. African Americans increased from 4.1% to 13.5% in census block group#
G53003300017013, located in the Greenwood neighborhood (Figure 12-A) and from
2.7% to 8.0% in census block group # G53003300021002, which covers parts of the
Meadowbrook and Wedgewood neighborhoods (Figure 12-B).
92
Figure 12 Positive change in African American population in gentrified census block
groups in various neighborhoods in Seattle from 2000 to 2010.
Gentrification trajectories varied for the census block groups that experienced
increases in their African American populations. Three census block groups that
gentrified in southwest Seattle led the transition with income above the Seattle
median in 2000 and experienced an increase in educational attainment above the
median by 2010. The census block groups that gentrified in north Seattle led with
above-median educational attainment in 2000 and followed with a transition to
above-median income by 2010. The census block group that experienced the greatest
positive change in the percentage of African Americans in Seattle in this period also
had the most significant gentrifying transition. This census block group
(#G53003300094004), located in the Beacon Hill neighborhood, was below the
median in all three categories in 2000 and transitioned to have all three categories
above the median in 2010 (Figure 12-D).
93
Asian populations decreased in 27 of the 43 census block groups that
gentrified. Decreasing values ranged from -49 % to -0.27%, with the majority of
gentrified census block groups ranging from -9% to 0% change in Asian population.
In some cases, when Asian populations decreased, African American populations
increased in census block groups that gentrified. This trend was most pronounced in
census block group # G53003300094004, located in the Beacon Hill neighborhood.
From 2000 to 2010, Asian population decreased in this census block group from 61
percent to 12 percent of the population (Figure 13-B); this is the census block group
in which African Americans increased dramatically over the same period (Figure 12D).
One other gentrified census block groups with a high percentages of Asian
population loss was located in North Seattle in Northgate neighborhood. This census
block group declined from 11.6 percent in 2000 to 1.5 percent in 2010. (Figure 13-A).
Figure 13 Spatial distribution of the greatest negative change in Asian racial
groups within gentrified census block groups from 2000 – 2010.
94
From 2000 to 2010, nine gentrified census block groups had a positive change
between 1 percent and 10 percent, they were located in North Seattle, Downtown,
Central Area and Capitol Hill and West Seattle neighborhoods. Of all the gentrified
census block groups two census block groups located in the Downtown and Cascade
neighborhoods had the greatest positive change. Census block group
#G53003300080015 located in the Downtown neighborhood increased. This census
block group was below the median for all three characteristics in 2000, by 2010
educational attainment and income were above the median. Census block group #
G53003300073003 increased from 10 percent to 21 percent from 2000 to 2010
(Figure 10). The census block group was below the median for all three
characteristics in 2000; by 2010 educational attainment and income were above the
median.
Figure 14 Greatest positive change in Asian population in a gentrified census
block group in the Downtown and Cascade neighborhoods, Seattle, from 2000 to
2010.
95
4.6. Relationship of racial change to gentrification
From 1990 to 2000, African American racial change varied across Seattle. In
1990, the highest percentages of African Americans (>50%) were located in the
Central Area just east of downtown Seattle, and high percentages of African
Americans were also found in South Seattle (Figure 15). From 1990 to 2000, African
American populations both increased and decreased as a proportion of census block
group populations. The greatest decline (of 15 to 34 percentage points) occurred in
spatially clustered pattern in central Seattle. African Americans also declined in
south Seattle by 5 to 14 percentage points, with some areas experiencing declines of
more than 24 percentage points. Over the same period, African American
populations increased by more than 2 percentage points, and in some cases more than
11 percentage points, in the most northern, southwestern, and southern portions of
Seattle.
From 2000 to 2010, declines in African American population were more
widespread throughout Seattle than in 1990 to 2000. As in 1990 to 2000, the greatest
declines (of up to 40 percentage points) occurred in central Seattle. African
Americans continued to increase as a percent of total population in northern Seattle.
In southwest Seattle, African American population decreased slightly in areas where
it had increased in the previous decade. In south Seattle, the African American
population increased in some areas where it had decreased in the previous decade.
Gentrified census block groups appear be clustered in north, central, and
southwest Seattle in both periods (1990 to 2000 and 2000 to 2010). Census block
groups that experienced gentrification from 2000 to 2010 were adjacent to census
block groups that gentrified in the previous time period, 1990 – 2000 (Figure 2 and
3).
Gentrified census block groups were associated with both declining and
increasing African American populations, but the relationships between gentrification
and changes in African American populations differ among central, northern,
southern, and southwestern Seattle. From 1990 to 2000, gentrification was associated
with declining African American populations in central Seattle (A-1 in Figure 13) and
in one census block group in south Seattle (B-1 in Figure 13). Two gentrified census
96
block groups in north and west Seattle had increases in their African American
populations, while the rest experienced no change in African American populations
(Figure 16A). From 2000 to 2010, African American populations declined in
gentrified census block groups in central Seattle and north Seattle (Figure 16 B). In
central and north Seattle, some of the census block groups that had experienced
increased African American population from 1990 to 2000 gentrified and lost African
American populations from 2000 to 2010. Small increases in African American
populations were associated with gentrification in census block groups in southwest
Seattle.
In some cases, gentrification preceded the decline in African American
populations. In census block groups that gentrified from 1990 to 2000, located just
west of the Central Area in Capitol Hill (Box A-1, Figure 13), the African American
population decreased slowly in 1990 to 2000, but more rapidly in the following
decade.
In other cases, including much of central Seattle, changes in African
American population preceded gentrification. In central Seattle, African American
populations declined from 1990 to 2000 in census block groups which did not fit the
definition of gentrification in 1990 to 2000, but which became gentrified in 2000 to
2010. For example, African American population declined by 15 to 24 percentage
points in a census block group located in the Lower Queen Anne area from 1990 to
2000. In 2000 to 2010, this census block group gentrified, and the African American
population continued to decline (Figure 13, Box A-1 and A-2).
97
Figure 15 - Percent African Americans in 1990
98
Figure 16. Gentrified and declining census block groups, change in percentage of
African American residents within each census block group, and places where
gentrification was perceived (a) from 1990 – 2000 and (b) 2000 – 2010. Census
block groups designated as "gentrified" experienced a change from below to abovemedian status in two or more of three categories: education, income, and home value
over the period. Census block groups designated as declining experienced a change
from above to below-median status in two or more of three categories: education,
income, and home value over the period.
4.7. Within-neighborhood spatial and temporal patterns of gentrification and racial
change
A closer look at three selected areas in Seattle reveals the complexity and
multidimensional aspect of gentrification, with contrasting patterns in different areas
(Figure 17). In central eastern Seattle, six census block groups gentrified from 1990
to 2000, and African American populations declined in two of these (Box A-1, Figure
17). From 1990 to 2000 in central Seattle, seven census block groups fell behind in
99
terms of education, home value, and income, and African American populations
declined in three of these. African American population also declined, and
gentrification was perceived, in many census block groups where gentrification was
not identified based on changes in education, home value, and income (Box A-1,
Figure 17). In central eastern Seattle from 2000 to 2010, nine census block groups
gentrified, and African American populations declined in eight of these (Box A-2,
Figure 17). Over the same period, nine census block groups had declining levels of
education, home value, and income, and African American populations increased in
four of these.
An area in northwestern Seattle experienced contrasting patterns from those in
central Seattle (Boxes B-1 and B-2, Figure 17). In this area, ten census block groups
gentrified in terms of education, income, and home value from 1990 to 2000, but did
not experience declines in African Americans. Asian populations both increased and
decreased in these areas (Figures 16 and 17). Interestingly, from 1990 to 2000 the
percent of African American population increased in one census block group whose
neighboring census block groups experienced gentrification and little to no change in
their African American populations over the same period (Box B-1, Figure 17). In
the following decade, this census block group gentrified, and the African American
population declined (Box B-2). In general, percentages of African American
populations did not change in this area from 1990 to 2000, but they began to decline
in 2000 to 2010 (Box B-1 and B-2, Figure 17).
Patterns of change in southern Seattle differed from those in central or
northern Seattle over the study period (Boxes C-1 and C-2, Figure 17). In the upper
half of the Rainier Valley and Beacon Hill (Box C-1, Figure 17), one census block
group gentrified from 1990 to 2000 and African American percentages declined
sharply. Five census block groups fell behind in terms of education, home value, and
income, and African American percentages declined in two of these. African
American percentages also declined in a number of census block groups, and
gentrification was perceived, but gentrification (in terms of education, home value,
100
and income) was not detected. From 2000 to 2010, African American populations
increased in some census block groups, and decreased in others. One census block
group in the Beacon Hill area gentrified and experienced a large increase in African
American population (Figure 17, Box C-1 and C-2).
Figure 17. Expanded views of selected areas shown in Figure 16.
4.8 Content analysis on select neighborhoods of gentrified census block groups 1990
- 2010
The results of the content analysis on gentrifying neighborhoods revealed a
pattern of news sources and blogs about gentrification and its effects on communities
within Seattle. Media reports and blogs primarily discussed a few neighborhoods:
Central Area, Capitol Hill, Columbia City, and Mount Baker in the Rainier Valley,
101
Beacon Hill, and Belltown. Three major themes emerged from the analysis: shifts in
demographics, a sense of community (or lack thereof in some cases), and economic
and development booms.
4.8.1. Demographic Shifts
A number of articles and blogs reported that communities experiencing
gentrification were also experiencing major shifts in racial demographics. The
Central Area (also known as the Central District) experienced gentrification and
declining minority population from 1990 to 2010 (Figure 17). Content analysis
showed that gentrification in these census block groups has changed the lives of
people in the neighborhood. For example the Seattle Times reported, “[the] Central
Area is now home to fewer African Americans than at any time in the past 30 years”
(Seattle Times 2001, July 22). The Stranger reported that the Central Area’s
improved economic conditions are linked to “large movement[s] of Blacks out of the
neighborhood.” However, the movement of Blacks is not always due to outright
displacement. The pastor of Mt. Zion, Central District’s oldest Black church, wrote,
“some black families are moving out of the traditionally black Central District to
posher areas like Kirkland, Newport Hills, and Mercer Island; [whereas], “poorer
blacks – victims of rising property taxes and ballooning housing prices – are being
forced out of Seattle” (Campell 2000).
Some view the change occurring in the Central District as favorable, while others
see it as “fundamentally destroying the cultural heritage of the city” (Richards 2015).
A longtime resident and one of the Central District’s black business owners, Earl
Lancaster, who owns the historically black barbershop, Earl’s Cuts and Styles,
perceived positive elements of the change. Lancaster recently hired a Caucasian
barber due to the change in the color of the neighborhood. Lancaster says, “The
community is getting more diverse …[and] I wanted to reflect that. It’s not just a
black barbershop. It’s a men’s grooming place.” A patron of the barbershop and
longtime resident of the Central District disagreed, indicating that he, “liked [the]
neighborhood the way it used to be” (Hansen 2013).
102
Gentrification is also occasionally linked to violence in media reports. For
example, in 2001, longtime activist James C. Garrett attacked the Mayor of Seattle.
Residents of the areas reported that “Garrett’s act was fueled by a growing well of
frustration that stems from issues ranging from accountability to crime in the
neighborhood, to issues of gentrification" (Jamieson, Murakami, and Rutherford
2001).
Changing demographics also were linked to violent acts in Capitol Hill, a
neighborhood located just north of the Central Area. A number of census block
groups in this area gentrified from 1990 to 2010, and cultures have clashed. Capitol
Hill was at one time dominated by predominantly gay households. Gentrification
brought a differing sexual orientation of its newcomers, and tensions have risen
between young straight people and gay people. The Seattle Times reported that a
young transgender man walking home from a club was verbally and physically
attacked by two straight men (Romano 2015). Another news article highlights an
anti-gay arsonist who endangered nearly 750 people at a gay nightclub on New
Year’s Eve in 2014 (Pulkkinen 2014). These events have spurred the mayor of
Seattle to implement a task force aimed at investigating methods of improving safety
in the Capitol Hill neighborhood (Garner 2015).
4.8.2. Sense of Community (or lack thereof)
The idea of community and building networks between old and new residents
was a reoccurring theme in the news and neighborhood blogs. In Beacon Hill,
“When new businesses move, they often don’t try to understand the people
who are here, their way of communicating and acting toward one another,”
says Khazm, [whose] family has lived in Beacon Hill for generations. “They
aren’t focused on building communities, their interested in investing in
property” (Uitti 2015).
The process of gentrification also has affected the sense of community and
belonging in the Central District. The Union Street Business Association noted that
as a result of gentrification, the Central District has “lost a sense of belonging, a sense
103
of safety, and a sense of ownership” (Jones 2015). One person remarked, “I thought I
would never see the time when I am driving through and white folks would look at
me strangely. I’m like, ‘I grew up in this area, what are you looking at me for?' ”
(Pisacane 2015). The struggle to integrate the cultures of longtime residents and
newcomers affects many neighborhoods experiencing gentrification.
4.8.3. Economic and development booms
The economic prosperity and increased development occurring in gentrified
neighborhoods has direct negative effects on longtime residents. In 2014, tenants on
the verge of eviction in the Ballard neighborhood picketed to protest a planned
renovation that would raise rents from approximately $890/month to $1500/month in
the Lockhaven apartment complex (Sears, K. 2014). The Freemont neighborhood has
also experienced an economic boom since the 1990s, with new coffee shops and highpriced townhouses in the commercial core. One resident commented, “We now have
more espresso bars than antique stores” (Freeman, P. 1999). In the Central District an
article noted that development of condominiums and triplexes along 23rd Avenue and
Rainier Avenue have raised rents and taxes to levels that prior residents cannot afford
(Mudede, C. 2005).
5. Discussion
Displaying gentrified census block groups overlaid on racial change and
perceptions of gentrification suggests that dynamic processes are at work in Seattle.
Gentrification is not binary, but exists on a spectrum, with many states and forms, and
it affects people and places differently. The initial definition of gentrification in this
study was based only on changes in education, income, and home value. The
addition of other information, such as changes in minority population and perceptions
of gentrification, provided a more nuanced picture of change in Seattle. For example,
from 1990 to 2000, gentrification coincided with declines in the African American
population and with accounts of perceived gentrification in some areas. However,
gentrification also was reported in areas that did not fit the definition of gentrification
in terms of change in education, income, and home value, but had declines in African
104
American populations and were near census block groups that did gentrify based on
this definition. Moreover, areas that experienced changes in education, income, and
home value, but lacked racial change, were also perceived as gentrifying based on
media accounts. These findings demonstrate that gentrification is a multidimensional, spatially contagious, dynamic process.
This analysis also reveals that the effects of gentrification do not always lead
to displacement of minorities. In some cases, African American populations slightly
increased in census block groups that gentrified. This may suggest that some African
Americans are able to take part in the process of gentrification and are exercising a
type of upward mobility. In other cases, census block groups declined in terms of
education, home values, and income, and also had increases in the African American
population. This may indicate migration of African Americans from some areas to
other areas, where they may take advantage of declining home values and rents. The
same processes may be affecting Asian populations.
The relationship of gentrification to change in minority population was related
to the density or concentration of minority residents. Outside of Seattle’s historically
black neighborhood, the Central Area, gentrification appeared to precede declines in
African Americans. However, within the Central Area and adjacent neighborhoods
where clusters of higher proportions of African Americans reside, gentrification
appeared to both precede and follow gentrification.
5.1. Hypothesis 1: Gentrification is more likely in neighborhoods with relatively low
income, home value, and education, as well as high proportions of minorities.
We expected to find more frequent gentrification in census block groups that
were below the median for educational attainment, home value and income at the
beginning of the study. Instead, household income was above the median at the
beginning of the study in many (44 percent) census block groups that gentrified,
although home value and educational attainment were below the median. This was a
surprising finding, because census block groups below the median in all three
categories have been thought to be more vulnerable to gentrification. The process of
gentrification happens gradually, and as we found, can reverse direction. Results of
105
this study support the interpretation that census block groups that were below the
median with respect to income, home value, and education experienced incremental
transitions over several decades, in which one socioeconomic variable increased first,
followed by increases in the others. Thus, a transition from below the median for all
three socioeconomic variables, to above the median for one of these, may be an early
indicator of gentrification (Figure 6 and 7). In most census block groups that
gentrified, household income transitioned to above the median first, followed by
increases in education and home value; this suggests that increased income promotes
upward mobility in terms of educational achievement and ability to take part in home
ownership.
5.2. Hypothesis 2: Gentrification is more likely to occur in locations adjacent to
areas that gentrified in a previous time period.
We expected census block groups to gentrify in a clustered pattern throughout
South Seattle. We found that gentrification was clustered among census block groups
in West Seattle and in North Seattle. Prior studies have also found clustering of
gentrification in Seattle in these areas (White 2012). Clustered patterns of
gentrification may be related to Tobler’s first law, that everything is related, but
things closer together are more related. An area may gentrify because of spillover
from neighboring wealthier census block groups. From 1990 to 2000, isolated census
block groups that gentrified were often adjacent to census block groups that were
above the city median for education, home value, and income. This process of
gentrification may be related to Smith’s (1979, 1982) rent gap theory. According to
the rent gap theory, gentrification will occur when existing buildings in an area have
depreciated to the point that the capitalized rent is less than the potential rent for the
building’s highest and best use (Smith 1979). The rent gap theory, which was
proposed at the household scale, may apply at the census block group level in this
study. For this study, when a neighborhood depreciated (fell below the median) so
that rents were very low, the disinvestment may have created circumstances for
reinvestment that was extremely profitable, leading to gentrification.
106
Other factors, both internal and external, may have contributed to the clustered
pattern of gentrification. Internal factors might include political, economic, or
occupational forces in the particular census block group. For example, in some areas,
a greater number of building permits may be issued. Or, neighborhoods may have
higher levels of community organization to promote development. Regarding the
occupational forces, technology companies have played a big role in Seattle (ie:
Adobe in Freemont and Amazon in Beacon Hill and while Microsoft is not located in
a particular neighborhood it sends charter busses to the Capitol Hill and other
neighborhoods to shuttle its employees to work. Intrinsic environmental factors also
could lead to the area being a hotspot for new investment. For example, property
values increase with proximity to environmental amenities such as parks or water
bodies (e.g Figure 3). Moreover, enhancements in environmental amenities may lead
to unintentional increases in property value (Aygeman 2001). Environmental
gentrification is the process of investment in the natural and built environment that
leads to increased property values and rents, and often leads to displacement of lowincome residents (Banzhaf and Walsh 2004). This topic is addressed in Chapter 4.
5.3. Hypothesis 3: Increases in income, home value, and education coincide with
declines in proportions of minorities.
We expected percentages of minorities to significantly decrease when census
block groups gentrified. As expected, proportions of African American and Asian
populations declined in gentrified census block groups, especially those that had high
proportions of these minorities at the beginning of the study.
Decreases in African Americans that coincided with gentrification were most
striking in the Central Area, but may have been driven by historical factors beyond
the scope of this study. The Central District is Seattle’s oldest Black neighborhood,
with low income, home values, and educational attainment. The history of the
Central Area (also known as the Central District) dates back to the creation of the
Federal Housing Administration, established in 1934, and the post-World War II
housing boom. The Federal Housing Administration was established to encourage
home ownership within the middle-class. However, racist covenants and a practice
107
called "redlining" prohibited Blacks, Asians and Jews from obtaining loans to
purchase homes in certain neighborhoods after World War II (Jackson, 1985). In
2006, University of Washington students made public the contents of 126 residential
neighborhood "covenants" affecting thousands of properties throughout Seattle that
were intended to exclude blacks and other minorities from purchasing homes in
certain neighborhoods. For instance, several neighborhood covenants explicitly
stated that, “No Hebrew or…any person of the Ethiopian, Malay, or any Asiatic
Race," or “No person of African, Japanese, Chinese, or of any other Mongolian
descent” should occupy homes within the neighborhood (Piscacane 2015). The
practice of redlining forced Blacks and some Asians to reside in the Central District.
This history is important because it reveals that gentrification is responding to factors
that have occurred over many decades prior to the period of this study (1990 to 2010).
5.4. Hypothesis 4: Media accounts perceive gentrification in places that experience
declines in populations of minorities, but not in places that only experience
changes in education, home value, and income.
We expected that the media content analysis would corroborate the quantitative
spatial analysis of gentrification. Many gentrified census block groups identified by
quantitative analysis also were discussed in media articles about gentrification.
However, there was an overwhelming amount of coverage of the Central District and
Capitol Hill, but not much about north and west Seattle. Moreover, perceptions of
gentrification were also found in areas that were not identified as being gentrified in
terms of education, income and home value.
5.5. Methodological issues
There are many approaches to defining and examining gentrification. This study
defined gentrification across the entire landscape of the City of Seattle, based on
decadal-scale changes in education, income, and home value in census block groups.
This approach was designed to reveal how changes in socioeconomic variables were
related to changes in minorities and how changes in neighborhoods were described in
the media. Other forms of data not used in this study also could help to describe
108
gentrification and reveal its causes and consequences. For example, data that depicts
changes in commercial activity within census block groups may reveal dramatic
changes that may be related to gentrification. Changes in the built environment also
are integral to the process of gentrification (Hwang and Sampson 2014).
The content analysis in this study could be expanded to include other forms of
media, such as videos and photographs, to explore how gentrification is portrayed
among neighborhoods. A content analysis also could be enhanced by interviews with
longtime residents, newcomers, and community activists and leaders within select
neighborhoods.
6. Conclusion
This study examined gentrification in Seattle at the landscape scale over two
decades, using quantitative analysis of census data on education, home value, income,
and race, as well as qualitative analysis of media content. Overall, the results of this
study indicate that gentrification is a dynamic, spatially variable process. When
gentrification was defined in terms of change in education, home value, and incomes,
it often coincided with declines in minority populations and perceptions of
gentrification in the media. However, in some cases gentrification defined in terms of
changes in education, home value, and income was associated with modest increases
in African American populations. The process of gentrification also played out
gradually over time: people may perceive gentrification, or African-American
populations may change, before an area experiences changed in education, home
value, and income.
More work is needed to understand the causes of gentrification. This chapter
showed that other factors besides those examined in this study may influence
gentrification. Chapter 2 showed that park access is related to socioeconomic
variables. Therefore, Chapter 4 explores how gentrification is related to changes in
environmental amenities.
109
7. Literature Cited
Anderson E.1990. Streetwise: Race, Class, and Change in an Urban Community.
Chicago, IL: The University of Chicago Press.
Aygeman, J., 2001. Ethnic minorities in Britain: short change, systematic indifference
and sustainable development. J. Environ. Policy Plan. 3, 15–30.
Balk, G. 2014. Yuppie playgrounds? Where gentrification is hottest in Seattle. The
Seattle Times http://blogs.seattletimes.com/fyi-guy/2014/02/14/yuppie-playgroundswhere-gentrification-is-hottest-in-seattle/ (last accessed 05 March 2016).
Barton, M. 2016. An exploration of the importance of the strategy used to identify
gentrification. Urban Studies. 53(1): 92-111.
Banzhaf, H. and R. Walsh. 2004. Testing for environmental gentrification: Migratory
Responses to changes in environmental quality. Paper for AERE Workshop. Estes
Park, CO.
Boone, C.G., G.B. Buckley, J.M. Grove, C. Sister. 2009.Parks and people: an
environmental justice inquiry in Baltimore, Maryland. Annals of the Association of
American Geographers. 99: 767-87.
Campell, P. 2000. Can I Get a Witness? Mt. Zion's Flame-Throwing New Preacher
Has Been Telling Seattle's Black Middle Class Exactly What They Don't Want to
Hear. The Stranger. http://www.thestranger.com/seattle/can-i-get-awitness/Content?oid=3299 (last accessed 05 March 2016).
Chang, S.H. 2015. Confronting “the g word” in Columbia City. The Seattle Globalist.
http://www.seattleglobalist.com/2015/07/07/confronting-gentrification-columbia-cityangeline-apartment/39095 (last accessed 05 March 2016).
Curl, A. 2008. Another step for gentrification in Columbia city. Seattle Weekly News.
http://www.seattleweekly.com/2008-12-03/news/another-step-for-gentrification-incolumbia-city/ (last accessed 05 March 2016).
Davila, F. and Mayo J. 2001. The Central Area: Seattle's Changing Heart. The Seattle
Times
http://community.seattletimes.nwsource.com/archive/?date=20010722&slug=black22
m1 (last accessed 06 March 2016).
110
Dizon, K. 2001. Is it Fremont – or Anywhere?: Gentrification ruffles funky ‘Center of
the Universe.’ Seattle Pi http://www.seattlepi.com/local/article/Is-it-Fremont-orAnywhere-1053242.php (last accessed 05 March 2016).
Kiersz, A. 2013. The most gentrified cities in America. Business Insider
http://www.businessinsider.com/map-the-cities-with-the-most-gentrifiedneighborhoods-2013-11 (Last accessed 05March 2016).
Freeman, P. 1999. Funky Freemont grapples with growth, gentrification. Biz
Journal. http://www.bizjournals.com/seattle/stories/1999/12/13/focus3.html (last
accessed 05 March 2016).
Garner, G. 2015. Capitol Hill gentrification blamed for rise in anti-gay hate crimes in
Seattle. Out.com. http://www.out.com/news-opinion/2015/6/03/capitol-hillgentrification-blamed-rise-anti-gay-hate-crimes-seattle (last accessed 05 March
2016).
Glass, Ruth. 1964. Introduction. London: Aspects of change. London: Centre for
Urban Studies and MacGibbon & Kee.
Hansen, B 2013. Changing face of the Central District. The Seattle Times.
http://www.seattletimes.com/photo-video/photography/changing-face-of-the-centraldistrict/ (last accessed 05 March 2016).
Holsti, O.R. 1969. Content analysis for the social sciences and humanities. Reading,
Mass.: Addison –Wesley.
Hwang J. and Sampson R.J. 2014 Divergent pathways of gentrification: Racial
inequality and social order of renewal in Chicago neighborhoods. American
Sociological Review 79: 726-751.
Jackson, R. 2014. Young families of color de-gentrifying Seattle? The Seattle
Globalist. http://www.seattleglobalist.com/2014/12/29/young-families-of-color-degentrifying-south-seattle/31824 (last accessed 05 March 2016).
Jackson, K. 1985. Crabgrass Frontier: The Suburbanization of the United States. New
York: Oxford University Press.
Jamieson, R., Murakami, K. and Rutherford, E. 2001, Attack on mayor spotlights
tensions in the Central District. Seattle Pi.
http://www.seattlepi.com/local/article/Attack-on-mayor-spotlights-tensions-inCentral-1059247.php (last accessed 05 March 2016).
Jenson, R.R. 2008. Measurement and comparison of Leaf Area Index estimators
derived from satellite remote sensing techniques. International Journal of Remote
Sensing. 25 (20): 4251-65.
111
Jones, M. 2015. The heart of Seattle: Stories from 23rd & Union. The Whidbey
Institute. http://whidbeyinstitute.org/23rd-union/ (last accessed 05 March 2016).
Kendig, H. 1984. Gentrification in Australia. In Palen and London 1984, 285-53.
Maurrasse DJ 2006. Listening to Harlem: Gentrification, Community, and Business.
New York: Routledge.
McGee, H. 2008. Seattle’s Central District, 1990—2006: Integration or
Displacement?
The Urban Lawyer 39:167
Minnesota Population Center. National Historical Geographic Information System:
Version 2.0. Minneapolis, MN: University of Minnesota 2011
Mudede, C. 2005. Going Under: In gentrified Seattle, the "hood survives on a few
small islands. But not for long. The Stranger.
http://www.thestranger.com/seattle/going-under/Content?oid=20859 (last accessed
05 March 2016).
Pisacane, C. 2015. The gentrification of Seattle’s Central District. The Seattle Times.
http://seattlecrimes.com/the_gentrification_of_the_cd.html (last accessed 05 March
2016).
Pulkkinen, L. 2014. 10 years in prison for Seattle gay club arsonist. SeattlePi.com
http://www.seattlepi.com/local/article/Neighbours-arson-Cheap-whiskey-orintolerable-5659983.php (last accessed 05 March 2016).
Ready, T. 1999. The other Red Apple Media ignores the CD. The Stranger.
http://www.thestranger.com/seattle/the-other-red-apple/Content?oid=1245 (last
accessed 05 March 2016).
Romano, T. 2015. Cultures clash as gentrification engulfs Capitol Hill. The Seattle
Times. http://www.seattletimes.com/life/lifestyle/culture-clash-on-capitol-hill/ (last
accessed 05 March 2016).
Sieg, H., V.K. Smith, H.S. Banzhaf, and R. Walsh. 2004. Estimating the general
equilibrium benefits of large changes in spatially delineated public goods.
International Economic Review 45:4 1047-77.
Sears, K. 2014. Gentrification debate heats up from ousted Ballard Apartment
Tenants. Seattle Weekly News. http://www.seattleweekly.com/home/951326129/gentrification-debate-heats-up-from-newly (last accessed 05 March 2016).
112
Smith, N. 1979. Toward a theory of gentrification; a back to the city movement by
capital, not people. Journal of American Planning Association. 45: 538-48.
_____ 1982. Gentrification and uneven development. Economic Geography. 58: 13955.
_____ 1996. The new urban frontier: Gentrification and the revanchist city. London:
Routledge.
_____ 2002. New Globalism, new urbanism gentrification as a global strategy.
Antipode. 34 (3): 427-450.
Richards, K 2015. A last-ditch effort to preserve the heart of the central district. The
Stranger. http://www.thestranger.com/features/feature/2015/02/25/21777706/a-lastditch-effort-to-preserve-the-heart-of-the-central-district (last accessed 05 March
2016).
United States Census Bureau, 2009-2013. Quick Facts.
http://www.census.gov/quickfacts/table/PST045215/00 <Last accessed 05 March
2016>
Uitti, J. 2015. Beacon Hill Blues: 3 local artists ponder the g-word. Culture.
http://crosscut.com/2015/05/beacon-hill-blues-3-artists-on-the-impact-of-the-g-word/
(last accessed 05 March 2016).
Vigdor, J. 2002. Does gentrification harm the poor? Brookings-Wharton papers on
Urban Affairs, Brookings Institution.
West Seattle restaurants: Gauadalajara's last weekend in Faunterloy. 2016. West
Seattle blog. https://westseattleblog.wordpress.com/ (Last accessed 05 March 2016).
Williams, P. 1984. Gentrification in Britain and Europe. In Palen and London 1984,
205-34.
White, J.D. 2012. Obscured geographies of the Emerald City: a study on
gentrification in Seattle, WA. WWU Masters Thesis Collection. Paper 191.
113
Chapter 4:
Gentrification and changes in park access and investment in Seattle, 1990 to
2010
Abstract
This study examined the relationships between spatial patterns of change and
investment in urban parks and processes of gentrification over time in Seattle,
Washington. A spatial analysis was conducted using GIS to test relationships
between changes in socioeconomic characteristics and changes in parks from 1990 to
2010. Changes in access to three different park types (recreation, natural passive, and
multiuse natural) were compared to changes in socioeconomic characteristics from
1990 to 2000 and 2000 to 2010. Spatial data on community-initiated investment in
parks were also compared to spatial patterns of change in park access and
socioeconomic characteristics. Gentrification was associated with locations where
the number and acreage of recreational and natural passive parks increased from 1990
to 2000, and where high levels of park investment ($500,000 to $1,000,000 per
census block group) had occurred from 2000 to 2010. Hence, changes in park access
and investment parks influenced the character of neighborhoods over time. However,
moderate and low levels of park investment were less associated with gentrification.
These findings contribute to understanding of resilience of neighborhoods and to
urban planning, because they indicate that there are thresholds of investment in park
improvement that may increase the likelihood of environmental gentrification, or help
a neighborhood to avoid environmental gentrification.
Keywords: GIS, parks, gentrification, environmental gentrification, community
investment
114
1. Introduction and Research Questions
Striking relationships between natural and built amenities and community
development have led researchers to investigate the intersections of race,
environmental quality, quality of life and the migration of residents. Because
environmental amenities such as parks can have positive effects on the quality of life
for urban residents, research has focused on the how these amenities can lead to
unequal distribution of their benefits through population displacement processes such
as gentrification or environmental gentrification.
Environmental gentrification occurs when environmental policies aimed to
improve quality of life primarily benefit higher income households (Sieg et al. 2004).
Improvements in environmental quality may initiate the process of environmental
gentrification, whereby high income people move in to a neighborhood with
perceived increasing environmental amenities, and property values and rents increase.
These changes may force poorer people to move to areas with lower environmental
amenities, where housing values are low (Banzhaf and McMormick, 2007).
Environmental gentrification has been shown to disproportionately affect minorities.
Revitalization of brownfields led to the decrease in Black and Latino populations in
various EPA classified regions across the U.S.(Essoka 2010). Case studies in
Vancouver and Toronto showed an inverse relationship between the investment in
greenery and the affordability of homes in the neighborhood (Dale and Newman
2009). Hence, investment in parks, or expansion of parks, may lead to environmental
gentrification.
This study took advantage of spatially and temporally explicit data on park types,
park investment and demographic data to test how the process of gentrification was
related to changes in environmental amenities, focusing on the park system, from
1990 to 2010 in Seattle, Washington, USA. We asked the following questions:
1. How were investments in parks related to changes in access to parks?
2. How were investments in parks related to gentrification?
115
3. How were changes in park access and park investments related to gentrification?
Hypotheses:
1. Park investment is spatially associated with increases in income, home
value, and education. If funding for park investment is based on proposals from
neighborhoods, those neighborhoods with more highly educated people may be better
equipped to submit successful proposals for park improvements. Also, home values
may increase because of a perception of improvement in neighborhoods. Therefore,
we expect that relatively high investment in parks will be spatially associated with
increases in home value, income, and education.
2. Increases in numbers or area of parks leads to increases in park investment.
The amenity value of parks depends on their number, area, and the quality of
facilities. The amenity value of parks can be increased by increases in park numbers,
park area, and investment in park facilities. Investments may be directed at
maintaining or renewing existing parks, or improving infrastructure in new parks. As
new parks are added to the park system, they attract additional investment.
Therefore, we expect that increased investment will occur in locations where park
numbers or park area increased over the previous decade.
3. Increased park access and investment in parks is spatially associated with
increases in income, home value, and education. Although Seattle has large areas of
lakeshore and coast, which provide environmental amenities, parks also are important
environmental amenities in Seattle. From 1990 to 2010, census block groups with
below-median levels of income and home value had access to significantly less
natural passive and multiuse parks, but more recreation parks, compared to census
block groups that were above the median. These patterns of association did not
change significantly from 1990 to 2010. Therefore, we expect that increased park
numbers and park area were spatially associated with increases in income, home
value, and education.
116
2. Study Site
Seattle, Washington, is the largest city in the Pacific Northwest with an area of
217 sq km and a population of 608,255 (US Census, 2010). The racial fabric of the
city is mostly Whites (69.5 %), Asians (13.8 %), and Blacks (7.9 %). American
Indian and Alaska Native, Native Hawaiian and other Pacific Islanders, and two or
more races together make up the remaining 8.8%. In contrast to the state of
Washington as a whole, Seattle is highly educated. As of 2010, 57.4 % of the
population age 25 and older held a bachelor’s degree or higher, compared to only
31.9% in the state of Washington and 28.8% in the US (United States Census Bureau,
2009-2013). However, less than half of the population of Seattle (46.8%) owns a
home, compared to 63.2% in the state of Washington and 64.9% in the US. The
proportion of households below the poverty line is 13.6%, which is comparable to the
rest of the state of Washington.
The selection of this study site was based on two key factors:
(3)
Seattle’s extensive gentrified history should be explored in a spatial
context. There is a great deal of media attention surrounding this city and
its underlying social processes. Seattle’s Central District transformed from
predominately working class African American neighborhoods to mostly
White, Asian American, and African American professionals, via
gentrification (McGee 2008). More recently, Seattle was featured as being
one of the “most gentrified cities in America” (Business Insider.com). In
February 2014, The Seattle Times reported that “two masked men blocked
a Microsoft employee shuttle on Capitol Hill…condemning the
neighborhood’s transformation into an ‘upscale yuppie
playground’…[with] skyrocketing rents that are displacing less-affluent,
longtime residents” (The Seattle Times 2014).
(4)
Seattle’s park system dates back to 1884, and consistent city investment
has made it leading park system in the world and an ideal place to explore
environmental gentrification. Seattle parks are funded in part from Seattle
residents, who have strongly supported park development. In recent years
117
voters three levies: the Community Center Levy ($36 million, 1999), the
Pro Parks Levy $198 million, 2000), and the Parks and Green Spaces Levy
($145 million, 2008). In addition, the Seattle Department of Parks and
Recreation also sought federal, state, and regional funding (Seattle Parks
and Recreation, 2014) and has defined a goal of accessibility, to ensure
that all residents live within 800m of a park (Harnik 2004).
3. Data sources and Methods
This study descriptive analysis to examine the spatial and temporal patterns of
park access and their relationships to socioeconomic variables and trajectories of
gentrification. The spatial distribution of park access was described based on changes
in numbers, area, and investment in park types within 800m of gentrified, declined,
and no change census block groups from 1990 to 2000 and 2000 to 2010.
3.1. Socioeconomic data sources
This study used socioeconomic data for the 1990, 2000, and 2010 US census.
The data were acquired from the National Historical Geographic Information System
through the Minnesota population center, University of Minnesota.
The geographic unit of analysis for all census related data was the census
block group. A block group is a statistical subset of a census tract, and generally
contains between 600 and 3,000 people (US Census Bureau 2012). Each block group
is uniquely identified allowing it to be represented graphically with a GIS. As of
2010, Seattle had a total of 481 block groups (Figure 1). In 2000 there were 570
census block groups, and in 1990 there were 630. Due to the changing boundaries of
census block groups over time, this study aggregated the 1990 and 2000 census block
group data to the 2010 census block group area based on an area weighting
methodology, which is explained later in this chapter.
Three socioeconomic variables were selected for this study (Table 2).
Variables were based on previous research in the quality of life, environmental
118
justice, environmental gentrification and environmental justice literature (e.g. Jensen
2008, Sieg et al. 2004 and Boone et al 2009). Median household income and median
home value were used to measure the economic status of the population. The
percentage of residents 25 and older with bachelor degrees or higher was used as a
measure of education (Table 2). In addition, racial groups were considered, including
Whites and minorities (Blacks and Asians).
The socioeconomic census data in .csv files and census block groups in
shapefile format were downloaded from the National Historical Geographic
Information System website. Using ArcGIS10.2.2, the .csv files and shapefiles were
imported into the ArcMap data frame. Using the join function, the .csv files and the
shapefiles were joined together using their unique identifying object ID number,
allowing each census block group to receive its corresponding socioeconomic
value.
Table 1. Socioeconomic variables collected by the US Census Bureau (Census
Bureau, 1990, 2000, 2010). These variables were used to analyze the process of
socioeconomic change from 1990 – 2010.
3.2. Data on park investment.
This study used the city of Seattle neighborhood matching funds database to
assess the amount of park investment within 800m of each census block group from
1990 to 2010. Seattle’s Department of Neighborhoods created the Neighborhood
Matching Fund (NMF) program in 1988 to encourage community participation by
119
providing the financial resources to create small and large projects. Projects can be
directed toward building community bonds through educational, cultural, or
recreational activities and projects. There are three types of project funds: (1) small
parks fund, (2) small and simple projects fund, and (3) large projects fund. Each fund
awards community organizations up to $10,000, $25,000 and $100,000, respectively
(Seattle Department of Neighborhoods 1995). Although this type of investment is
only one of several sources of investment in parks, this dataset was readily available
in a useable format for this analysis.
The NMF dataset was obtained from the city of Seattle data portal. The NMF
database provides documentation for all awarded neighborhood matching funds from
1989 to present. Information associated with this database include: ID number,
Contract number, Program Area (i.e. project fund type), award year, project title,
organization name, project description, amount awarded, pledged amount, and the
neighborhood being served. The dataset was not available in a form that could be
effectively used in GIS or for the purposes of this study. Therefore, the information
was processed and filtered to identify which records were associated with park
improvements. A Python code was written to sort through the database to find certain
park names and projects associated with the study. Once the park name was found, it
was written to another excel file. Then using a GIS the new file was joined with the
parks shapefiles based on their unique park name. The total amount of the awards
was summarized for each park from 1990 to 2000 and 2000 to 2010. To identify
which census block groups received funding for parks within 800m, an 800m buffer
was placed around each census block group. The park shapefiles were then overlaid
on the buffer layer and spatially joined together using a one to many join operation
and an intersect match option. The output was a shapefile in which the attributes
from one shapefile were joined to the other shapefile based on their spatial
relationship. The attribute table and total award money was summarized based on the
unique census block group ID to assess the total amount of money that was awarded
to parks within 800m of each census block group. After the data was processed and
joined with the spatial datasets further analysis could occur.
120
3.3. Determine changes in park access.
Park access was defined as the number and area of parks within a census block
group and within 800 m of the census block group. Changes in park access from
1990 to 2000 and 2000 to 2010 were determined by differencing the fields associated
with the number or area of parks within 800m in 2000 from the 1990 field. This was
repeated for the 2010 and 2000 fields. Spatial data on changes in park numbers and
area for each census block group were obtained from Ch. 2 of this dissertation.
3.4. Identify gentrified census block groups
Gentrification was measured at the census block group level and defined as an
increase from below to above-median values for home value, income, and education
for periods of 1990 to 2000 and 2000 to 2010. Declining census block groups were
defined as census block groups that transitioned from above to below-median in at
least two or more categories. Spatial data on gentrified and declining census block
groups were obtained from Ch. 3 of this dissertation.
3.5. Define environmental gentrification and how environmental gentrification will be
tested.
Environmental gentrification was defined as the spatial association between park
investment, increases in park access, and gentrification. GIS layers representing park
investment, changes in park access, and gentrification were joined and overlaid.
Spatial associations were identified by visual inspection of the resulting maps.
4. Results
4.1. Park investment vs. gentrification.
The number of census block groups which received park investment from
1990 to 2000 was highest for recreation park type and lowest for natural passive park
121
type (Figure 1a). From 2000 to 2010, the number of census block groups which
received park investment was highest for recreation park type and lowest for natural
passive park type (Figure 1b). From 1990 to 2000, the average investment for a
census block group with parks within 800m was highest for recreation parks, and
lowest for natural passive parks (Figure 2a). From 2000 to 2010, average investment
for a census block group with parks within 800m was highest for recreation parks,
and lowest for natural passive parks (Figure 2b).
From 1990 to 2000, investments in recreation and multiuse parks were made
in slightly fewer census block groups that declined compared to those that gentrified
or did not change (Figure 1a). From 2000 to 2010, for census block groups that
declined and recreation or multiuse parks within 800m fewer investments were made
compared to those census block groups that gentrified or did not change. (Figure 1b).
From 1990 to 2000, investments in natural passive and multiuse parks averaged only
about $30,000 in declining census block groups but were approximately $175,000 in
census block groups that gentrified (Figure 2b). From 2000 to 2010, investments in
all parks increased for declining census block groups averaging approximately
$80,000 for natural passive parks, $180,000 for recreation parks, and $200,000 for
multiuse parks. Gentrified census block groups averaged the highest investment
values for recreation and natural passive parks but the least in multiuse parks
compared to census block groups that did not change or declined (Figure 2b).
122
Figure 1 Percent of census blocks in this category, in which there was some (nonzero)
investment in parks from (a) 1990-2000. and (b) 2000-2010.
123
Figure 2 Average size of park investment for census block groups which gentrified,
did not change, or declined from (a) 1990-2000 and (b) 2000-2010.
124
4.2. Changes in park access and gentrification.
4.2.1. Changes in park access and gentrification - recreation parks
From 1990 to 2000, census block groups gained up to six recreation parks
within 800 m (Figure 3). Most of these increases occurred in central and south
Seattle. Census block groups located outside of these areas experienced little to no
change in the number of recreation parks within 800m. One gentrified census block
group in South Seattle gained four recreation parks within 800m. Census block
groups that declined were associated with little to no change in the number of
recreation parks within 800m.
From 2000 to 2010, census block groups also gained up to six recreation parks
within 800 m. Two distinct hotspots of change in the number of recreation parks are
evident. These hotspots are located in North Seattle and at the nexus of four
neighborhoods in South Seattle including Downtown, the Industrial District,
International District, and North Beacon Hill. Two census block groups in North
Seattle gained five recreation parks from 2000 to 2010, and gentrified during the
same period. One census block group in the South Seattle hotspot gained three
recreation parks within 800m from 2000 to 2010, and also gentrified. Five gentrified
census block groups in West and North Seattle gained two recreation parks within
800m. The remaining gentrified census block groups had little to no change in the
number of recreation parks within 800m from 2000 to 2010 (Figure 3).
125
Figure 3 The change in the number of recreation parks within 800m of each census
block group compared to gentrified and declined census block groups from 1990 to
2000 and 2000 to 2010.
From 1990 – 2000, most increases in recreation park acreage occurred in four
hotspots located in South and North Seattle (Figure 4). The greatest increase in
acreage of recreation parks (25 acres per census block group) occurred in South
Seattle, in the Central Area and Rainier Valley neighborhoods. In the Downtown
area, census block groups gained 20 to 25 acres of recreation park land within 800m.
In North Seattle, census block groups in two areas gained 15 to 20 acres of recreation
park. However, only one census block group in South Seattle gained more than 25
acres of recreation park from 1990 to 2000 and also gentrified. Another census block
group in North Seattle gained 15-20 acres of recreation park and also gentrified.
However, other gentrified census block groups had little to no change in recreation
126
park area. Most census block groups that declined had little to no change in
recreation park acreage. However, one census block group in North Seattle gained
15-20 acres of recreation park but declined, and a cluster of census block groups in
Downtown Seattle gained 15-20 acres of recreation park from 1990 to 2000, but
declined over that period.
From 2000 to 2010, recreation park acreage did not change in most census block
groups. A cluster of census block groups just South of Union Lake and another
cluster in North Seattle gained 5-10 acres of recreation park within 800m from 2000
to 2010. Most gentrified and declined census block groups had little to no change
recreation park acreage from 2000 to 2010 (Figure 4).
Figure 4 The change in the acreage of recreation parks from 1990 – 2000 and 2000 –
2010 in relation to gentrified and declined census block groups.
127
4.2.2. Changes in park access and gentrification - natural passive parks
From 1990 to 2000 census block groups gained up to four natural passive
parks within 800m (Figure 5). Census block groups that gained three or more natural
passive parks were mostly associated with declined census block groups or census
block groups that did not change with respect to gentrification. Approximately
twelve gentrified census block groups gain one or more natural passive parks within
800 m from 1990 to 2000. Most of these gentrified census block groups gained only
one additional natural passive park, but one gentrified census block group in the
Fremont neighborhood gained two new natural passive parks, and two of the
gentrified census block groups gained three natural passive parks within 800 m one
(in the Central Area the Queen Anne neighborhood) (Figure 5).
From 2000 to 2010, fewer natural passive parks were added within 800m of
most census block groups (Figure 5). Most census block groups did not gain any
natural passive parks within 800m. Census block groups that gained two or more
parks occurred in North (Northgate and University District), Central District, Rainier
Valley and West Seattle. Two census block groups in the Industrial District (adjacent
to West Seattle) that gentrified from 2000 -2010 gained three parks in the previous
decade (1990 to 2010). In West Seattle, three census block groups that declined from
1990 to 2000 and gained 2 to 3 new parks also gentrified from 2000 to 2010. The
same trend is evident in Northeast Seattle, where two census block groups declined
from 1990 to 200, but gained two to three new parks, and gentrified from 2000 to
2010 (Figure 5).
128
Figure 5 The change in number of natural passive parks within 800m of each census
block group from 1990 to 2000 and 2000 to 2010, compared with gentrified and
declined census block groups during the same time periods.
From 1990 to 2000, the majority of census block groups had little to no change in
natural passive park acreage within 800m (Figure 6). Census block groups in West
Seattle in the Delridge neighborhood gained 100 to 150 acres of natural passive park
land. Gentrified and declined census block groups did not appear to be associated
with changes in natural passive park acreage from 1990 to 2000.
From 2000 to 2010, the largest change in area of natural passive parks (354
acres within 800 m) occurred in one census block group in northwest Seattle. Other
census block groups in north Seattle gained 100 to 200 acres of natural passive park
land within 800m. Most gentrified census block groups had little or no change in
129
natural passive park area. However, three census block groups in North Seattle,
which declined from 1990 to 2000, gained 100 to 150 acres of natural passive park
area and gentrified from 2000 to 2010. Declined census block groups appeared to
have little to no change in natural passive park area from 2000 to 2010.
Figure 6 The change in the acreage of natural passive parks from 1990 to 2000 and
2000 to 2010 in relation to gentrified and declined census block groups.
4.2.3. Changes in park access and gentrification - multiuse parks
Very few multiuse parks were added within 800m of census block groups
from 1990 to 2010 (Figure 7). Three clusters of census block groups located in the
Northwest Seattle, Queen Anne, Industrial District, and Beacon Hill neighborhoods
130
gained at least one multiuse park within 800m. Change in numbers of multiuse parks
did not appear to be related to gentrified census block groups during this time period.
From 2000 to 2010, six clusters of census block groups gained one multiuse
park within 800m. The clusters were located in South Seattle (Delridge and Industrial
District neighborhoods), Central District, Queen Anne, Magnolia, and three clusters
were located in North Seattle. Ten of the 43 census block groups that gentrified from
2000 to 2010 also gained one multiuse park in that period. Four of these were located
in the Central Area neighborhood, two in the Queen Anne neighborhood just north of
Downtown, one in the Ballard neighborhood, two in the Fremont neighborhood just
north of Queen Anne and one in the Northgate neighborhood in North Seattle.
However, approximately 16 census block groups that declined from 2000 to 2010
were also associated with one new multiuse park within 800m during the same time
period (Figure 7).
131
Figure 7 The change in the number of multiuse parks within 800m of each census
block group compared to gentrified and declined census block groups.
From 1990 to 2000, although multiuse park acreage did not change within 800
m of most census block groups, a cluster of census block groups in North Seattle
gained more than 50 acres of multiuse park area (Figure 8). One census block group
adjacent to this area gentrified in the same period. However, multiuse park area did
not change within 800 m of the other gentrified and declined census block groups
during this period.
From 2000 to 2010, multiuse park acreage also did not change within 800 m
of most census block groups. Gentrified and declined census block groups did not
appear to be associated with changes in multiuse park acreage (Figure 8).
132
Figure 8 The change in the acreage of multiuse parks from 1990 – 2000 and 2000 –
2010 in relation to gentrified and declined census block groups.
4.3. Park investment, changes in park access, and gentrification.
4.3.1. Park investment, changes in park access and gentrification - recreation parks
Investments in recreation parks were larger from 2000 to 2010 than from 1990
to 2000 (Figure 1,2, and 9). From 1990 to 2000, census block groups in the
Northwest portion of Seattle gained one park within 800m and investment was low.
However, from 2000 to 2010, this area became gentrified, and more than $1,000,000
was invested into the recreation parks within 800m of the census block groups
(Figure 9). From 1990 to 2000, census block groups in South Seattle gained five or
six recreation parks within 800m, and investment in recreation parks was moderate
133
($300,000 to $500,000). However, by 2010, this area had slightly higher levels of
investment in recreation parks, and one census block group became gentrified. From
2000 to 2010, areas in Northeast and downtown Seattle gained five or six recreation
parks within 800m. Park investment in these areas was also high, and several census
block groups gentrified in this area over this time period (Figure 9).
Figure 9 Investment in recreation parks compared to the change in number of
recreation parks within 800m of each census block group from 1990 to 2000 and
2000 to 2010 in relation to gentrified and declined census block groups.
From 1990 to 2010, park acreage increased in two distinct areas, followed by
large investments in recreation parks from 2000 to 2010, which coincided with
gentrification in the same census block groups (Figure 10). In Northeast Seattle,
census block groups gained 15 to 20 acres of recreation parks within 800m and
134
$500,000 to $1,000,000 investments in parks from 1990 to 2000. In the subsequent
decade, recreation park investments fell to less than $100,000 and most of the area
declined from 2000 to 2010. In downtown Seattle, census block groups gained 20 to
25 acres of recreation parks within 800m from 1990 to 2000. In the next decade,
recreation park investments in this area exceeded $1,000,000, and the area gentrified.
In contrast, census block groups that gentrified in portions of South Seattle and
gained more than 25 acres of recreation parks within 800m from 1990 to 2000,
declined from 2000 to 2010 but park investment slightly increased (Figure 10).
Figure 10 Investment in recreation parks compared to the change in acreage of
recreation parks within 800m of each census block group from 1990 to 2000 and
2000 to 2010 in relation to gentrified and declined census block groups.
135
4.3.2. Park investment, changes in park access and gentrification - natural passive
parks
From 1990 to 2010, the number of natural passive parks increased in three
distinct clusters of census block groups, and large park investments and gentrification
occurred in the next decade (Figure 11). From 1990 to 2000, in Northwest Seattle, a
cluster of census block groups gained two natural passive parks, and less than
$10,000 was invested in these parks. However, from 2000 to 2010, investments in
the natural passive parks within 800m of these census block groups increased to
$500,000 to $1,000,000, and the census block groups gentrified. Similar patterns
occurred in the Queen Anne (just west of Lake Union) and West Seattle
neighborhoods. Census block groups in these neighborhoods gained four natural
passive parks within 800m from 1990 to 2000. Park investment in census block
groups in these neighborhoods was moderate ($10,000 to $100,000 per census block
group) from 1990 to 2000. In the following decade, investment in natural passive
parks in these neighborhoods increased to $500,000 to $1,000,000 per census block
group, and some census block groups gentrified (Figure 11).
136
Figure 11 Natural passive park investment compared to the change in number of
natural passive parks within 800m of each census block group from 1990 to 2000 and
2000 to 2010 in relation to gentrified and declined census block groups.
From 1990 to 2000, each census block group in the Delridge (just east of West
Seattle neighborhood) area gained 100 to 150 acres of natural passive parks within
800m (Figure 12). In the following decade, investments in natural passive parks
increased to $500,000 and $1,000,000, and census block groups gentrified in this
area. In northwest Seattle, large increases in natural passive park area coincided with
large investments in these parks and gentrification from 2000 to 2010 (Figure 12).
137
Figure 12 Investment in natural passive parks compared to the change in acreage of
natural passive parks within 800m of each census block group from 1990 to 2000 and
2000 to 2010 in relation to gentrified and declined census block groups.
4.3.3. Park investment, changes in park access and gentrification - multiuse parks
Overall park investment was not obviously associated with changes in the
number of multiuse natural parks (Figure 13). Most investment in multiuse parks
occurred in census block groups located along the Puget Sound and Lake Washington
and in North Seattle (Figure 13). From 1990 to 2000, the highest investment in
multiuse parks ($500,000 to $1,000,000 per census block group) occurred in South
Seattle, but this area did not gentrify in the following decade; in fact, census block
groups declined. However, in other parts of Seattle, census block groups that
declined had small investments (up to $100,000 per census block group) and little to
138
no change in the number of multiuse parks from 1990 to 2000. Gentrified census
block groups had a wide range of investment in multiuse parks from 1990 to 2000.
One gentrified census block group in the lower Central District had between
$500,000 and $1,000,000 in multiuse park investment within 800m. The following
decade this census block group declined but park investment increased to more than
$1,000,000 (Figure 13).
From 2000 to 2010, investment in multiuse parks exceeded $1,000,000 in six
census block groups, which were all located in South Seattle. The smallest
investments for multiuse parks within 800m of census block groups appear to be
dispersed uniformly across North and West Seattle. Gentrified census block groups
appeared to be associated with all levels of park investment. Gentrified census block
groups that had high levels of multiuse park investment were found in Central
District, Queen Anne1, Magnolia2, and Ballard3.
139
2
1
1
3
1
Figure 13 shows the multiuse park investment compared to the change in number of
multiuse parks within 800m of each census block group from 1990 to 2000 and 2000
to 2010 in relation to gentrified and declined census block groups.
In most cases, increases in multiuse park acreage were not associated with
large investments (Figure 14). However, in some cases, census block groups
gentrified and also had high levels of park investment within 800m. The Northeast
corner of Seattle gained the largest amount of multiuse park acreage from 1990 to
2000 (81 acres), and one census block group adjacent to this area gentrified. From
2000 to 2010, four census block groups in the Magnolia1 neighborhood gained 10 to
30 acres of multiuse park area, and this coincided with park investments of $500,000
to $1,000,000. One of these census block groups gentrified during this time period.
Some census block groups along Lake Washington in South Seattle had little to no
140
change in multiuse park acreage but received more than $1,000,000 in investments in
multiuse parks.
1
Figure 14 Investment in multiuse natural parks compared to the change in acreage of
multiuse natural parks within 800m of each census block group from 1990 to 2000
and 2000 to 2010 in relation to gentrified and declined census block groups.
5. Discussion
Gentrification, measured as an increase from below to above-median levels of
home value, income, and education, was expected to be associated with increases in
park numbers and park area in the same time period. However, the results were not
consistent with this expectation: only two census block groups with very large
changes in the acreage of recreation parks from 1990 to 2000 and natural passive
141
parks from 2000 to 2010 also gentrified in the same period. I found that numbers and
acreage of parks do not predict gentrification, but the addition of information on park
investment reveals strong lagged relationships. In census block groups where
numbers and areas of parks increased, followed by high investments in parks, led to
gentrification. In the Downtown and South Lake Union neighborhoods, acreage of
recreation parks increased from 1990 to 2000, moderate ($300,000 to $500,000) to
high ($500,000 to $1,000,000) park investments occurred from 2000 to 2010, and
these areas gentrified.
Traditional trajectories of gentrification suggest that growth in professional
and business services coupled with high technology jobs would lead to gentrification
in Seattle (Morrill 2008). However, the trends found in this study are indicative of
processes that are far more complex than the traditional notions of gentrification as
merely changes in socioeconomic status. The results suggest that changes in access
to, and investment in, differing park types (particularly natural passive and recreation
parks) may play a significant role in explaining social, economic, cultural, and in
some cases, racial changes to census block groups in Seattle, Washington.
The environmental gentrification literature is traditionally associated with
environmental hazard cleanup sites that relate to the increased attraction of the area
and ultimately lead to gentrification (Banzhaf and McCormick 2007, Eckerd 2011).
The processes described in this study fit the definition of environmental gentrification
because improvements in the landscape (i.e., increased numbers and area) appeared to
catalyze the process of gentrification in the following decade and led to the
displacement of vulnerable populations. Abel et al (2015) argued that gentrification
accompanied and at times preceded environmental improvements in Seattle.
However, this study assessed environmental gentrification in terms of park access and
investment, and provides new evidence that spatial patterns of gentrification can be
attributed to changes in the numbers, areas, and investment in parks in Seattle.
This study examined the relationships of changes in three different types of
parks to trajectories of gentrification. Some research has suggested that the causes of
gentrification are well understood. However, this study shows that gentrification is
142
far too complex to attribute to simple factors, and that the causes of gentrification are
multifaceted, dynamic, and spatially variable. Although environmental gentrification
is occurring in Seattle at the census block group level, the patterns and processes
differ among neighborhoods. Changes in certain types of parks may be more related
to gentrification than others. High levels of park investment at the neighborhood
level also appear to be a strong predictor of gentrification. High-income, relatively
more educated community groups may be more organized and more successful in
competing for funding for parks.
These findings also suggest that moderate investment in parks may not
necessarily lead to gentrification. This is a strategy called, "just green enough." The
"just green enough" strategy is based on the idea that communities can enhance their
parks by small improvements, without spurring the process of gentrification (Curran
and Hamilton 2012). If communities restrict park improvement projects to moderate
levels and ensure that they are designed by and for the community, park investment
may avoid gentrification.
6. Conclusion
Gentrification is a multifaceted, dynamic and spatially variable process. Research
on the causes of gentrification is transcending early theories of cultural and economic
changes to examine the role of environmental improvements. Increases in park
access, when combined with high levels of investment in parks, appeared to explain
spatial patterns of environmental gentrification in Seattle. The process of
environmental gentrification also seems to be sensitive to changes in different park
types. These findings contribute to the environmental justice because they reveal
persistent inequities in the distribution of environmental goods in Seattle. More
research is needed to explore relationships of park access and investment and their
relationship to gentrification and environmental justice in Seattle.
143
7. Literature Cited
Abel, T.D., J. White, and S. Clausen. 2015. Risky Business: Sustainability and
industrial land use across Seattle’s gentrifying riskscape. Sustainability. 7(11): 1571815753.
Banzhaf, H., and E. McMormick. 2007. Moving beyond cleanup: Identifying the
cruicbles of environmental gentrification. Working paper 07-29, Andrew Young
School of Policy Studies.
Boone, C.G., G.B. Buckley, J.M. Grove, C. Sister. 2009.Parks and people: an
environmental justice inquiry in Baltimore, Maryland. Annals of the Association of
American Geographers. 99: 767-87.
City of Seattle Neighborhood Matching Funds database. City of Seattle. 2016.
https://data.seattle.gov/browse?category=Community&limitTo=datasets&utf8=%E2
%9C%93 (last accessed 07 March 2016).
Curran W. and T. Hamilton. 2012. Just green enough: contesting environmental
gentrification in Greenpoint, Brooklyn. Local Environment. 17 (9): 1027-1042.
Dale, A. and L.L Newman. 2009. Sustainable development for some: green urban
development and affordability. Local Environment: The International Journal of
Justice and Sustainability, 14 (7): 669–681.
Eckerd, A. 2011. Cleaning up without clearing out? A spatial assessment of
environmental gentrification. Urban Affairs Review. 47 (1): 31-59.
Essoka, J.D. 2010. The gentrifying effects of brownfields redevelopment. Western
Journal of Black Studies 34 (3): 299–315.
Harnik, P. 2004. Parks: How far is too far? Planning. 9-11.
Holling, C.S. 1973. Resilience and stability of ecological systems. Annual Review of
Ecology and Systematics. 4: 1-23.
Jenson, R.R. 2008. Measurement and comparison of Leaf Area Index estimators
derived from satellite remote sensing techniques. International Journal of Remote
Sensing. 25 (20): 4251-65.
McGee, H. 2008. Seattle’s Central District, 1990—2006: Integration or
Displacement? The Urban Lawyer 39:167.
Morrill, R. 2008. Growth Management, the Market, and Settlement Change in Greater
Seattle 1990-2007 https://faculty.washington.edu/morrill/papers/GMWRSA.PDF (last
accessed 07 March 2016).
144
Minnesota Population Center. National Historical Geographic Information System:
Version 2.0. Minneapolis, MN: University of Minnesota 2011.
Neighborhood Matching Funds. Seattle Department of Neighborhoods 1995.
http://www.seattle.gov/neighborhoods/programs-and-services/neighborhoodmatching-fund (Last accessed 06 March 2016).
Sieg, H., V.K. Smith, H.S. Banzhaf, and R. Walsh. 2004. Estimating the general
equilibrium benefits of large changes in spatially delineated public goods.
International Economic Review 45:4 1047-77.
United States Census Bureau, 2009-2013. Quick Facts.
http://www.census.gov/quickfacts/table/PST045215/00 <Last accessed 05 March
2016>
Walker, B. and Salt, D. 2006. Resilience Thinking. Sustaining ecosystems and people
in a changing world. Washington DC: Island Press.
145
Chapter 5 - Conclusion
In this study I examined the relationships among different factors including
wealth (household income and home value) and education, race, perceptions of
gentrification park access and investment in parks. In short I sought out to
understand how these factors were related to the spatial patterns of gentrification
environmental justice and concepts of land use planning.
Revisiting the Society and Land Use conceptual model adapted from Platt,
will help to define the gains made by this research and understand where this research
needs to go in the future.
Figure 1. The modified Society and Land use model (adapted from Platt, 2004).
146
In Chapter 2 I examined the relationships among park access and development
of new parks by type, wealth, education, and racial patterns, in order to understand
how park system evolution is related to environmental justice. I found that park
access differed among socio-economic groups, and by park type. Significantly
greater proportions of minorities were found in census block groups with below –
median levels of education, income, and home value, which were related to
significantly larger numbers and more areas of recreation parks. I also found that
park development had the effect of maintaining differences in park access based on
socio-economic variables. The modified Society and Land Use model shows how the
socio-economic variables feed into the legal realm and inform the resource
management and the type of parks that are developed and where. Over time Seattle
has seen a large increase in the number and size of parks largely due Seattle voters
positively expressing the need for park space through the approval of bonds and
levies. However, this study found disparities in park access by park type and socioeconomic group. Understanding the historic development context of the Seattle park
system is an important factor that deserves more attention in the future. It invokes the
question: Is there a systemic approach that reinforces an uneven distribution of park
access? This portion of the research could be enhanced by interviewing key park
managers on the role that park planning has taken over the years and whether or not it
has targeted certain neighborhoods for development of particular park types. It is
important for community planners to recognize that there may be disparities in access
to parks based on park type.
In chapter 3 I examined gentrification over two decades, using quantitative
analysis of census data on education, home value, income, and race, as well as
qualitative analysis of media content. This chapter focuses on the characteristics and
trends of the economic and social data vector of the Society and Land use Model
(Figure 1). It not only characterizes the economic and social data vector but breaks it
down to reveal its own dynamic process. I found gentrification to be a dynamic and
variable process that exists on a continuum rather than being a binary process. I
identified three trajectories of gentrification. Gentrification was defined as change
from below to above the median value for education, income, and home value in a
147
census block group over the period 1990 to 2000 or 2000 to 2010. Census block
groups showing this trend were signified as Gentrification A. Nine percent of census
block groups met the definition of gentrification in each of the two decades. Some of
the gentrified census block groups were also accompanied by large declines in
proportions of African Americans in the population. However, gentrification in terms
of education, income, and home value also occurred in areas that did not have large
proportions of minorities. The interaction of wealth and education as it related to
racial change and gentrification was another trajectory identified as Gentrification B.
The last factor added was media perceptions of gentrification. We found that
perceptions of gentrification were reported in places that met the definition of
gentrification, but also in areas experiencing changes in minorities, but which did not
yet meet the definition of change in terms of education, home value, and income.
This interaction was identified as Gentrification C. I also found that Census block
groups gentrified in a clustered manner and Census block groups that gentrified from
2000 to 2010, tended to be in close proximity to other Census block groups that
gentrified in the previous decade from 1990 to 2000. This trend suggests that
gentrification may have a contagion effect.
Studying trajectories of gentrification in this manner adds to the dialogue of
change especially concerning how people perceive the change in their neighborhood
even when gentrification may not be detected. Further research on this dynamic
process presents a unique opportunity for community leaders and planners to identify
whether or not perceptions of gentrification are actually precursors to it being
detected in census data and other quantitative methods. In practice this is an
opportunity for communities to combat or brace themselves for impact. It may serve
as a window for local government to reinforce or target social programs and
initiatives to certain areas. Community leaders may use this as means of foreseeing
and sidelining the typical pitfalls that other gentrified communities have struggled
with. For example, the content analysis in this research identified that gentrification
stifled the community bonds and sense of togetherness. Therefore, community
initiatives may focus on strengthening the community so that it reinforces or
reestablishes its identity, allowing newcomers including businesses and other
148
gentrifiers to not change but enhance the already vibrant nature of the community.
Other initiatives may target long standing residents, both home owners and renters.
Local housing authority can work to ensure that affordable housing is maintained as
developers move into the communities. Gentrification is a multifaceted yet slow
process. The slow nature of gentrification can be a beneficial feature of the process
or act as a buffer to help individuals and the community at large to prepare for the
changes that may occur.
The research conducted in Chapter 3 was limited in two areas: (1) the time
frame being too short and (2) the content analysis focusing on the negative
perceptions of gentrification. As previously mentioned gentrification is a slow
process a two decade time frame is just a small window or snap shot in time which
makes it difficult to effectively draw conclusions and analyze trends. Future research
should provide a more robust timeline in order to gain a fuller picture of the evolution
that the city of Seattle has taken. Secondly, the approach to conduct the media
content analysis was limited due its keywords. The keywords used rendered an
overwhelming response of negative perceptions of gentrification. Gentrification for
many has been seen as a negative process due to its ability to displace low income
and minority persons. However, not everyone is of the same mindset and to some
gentrification is seen as a process that brings vitality back to their neighborhood and
is gladly welcomed. This research did not find particularly find perceptions of
gentrification even in areas where gentrification is a known process to be occurring.
This may have occurred because the community is not reporting about gentrification
or because the search terms, as mentioned were limited due to their negative
connotation. Further research should broaden search terms to include positive terms
of gentrification. Another means of creating a more robust research agenda of
perceptions is to conduct interviews within and surround the gentrified Census block
groups. Interviews at all levels of the community would be beneficial for the research
including; local business owners, renters, home owners, community leaders and
activists, as well as corporate business owners, and city officials communities.
149
Chapter 4 refocuses its attention on the interrelationships of the modified
Society and land use model. Spatial patterns of park access were compared to the
process of gentrification A. Spatial data on community-initiated investment in parks
were also compared to spatial patterns of change in park access and socio-economic
characteristics. Using the Society and Land Use model I explore the interactions
between the physical realm and changes in counts or area of park types, the economic
and social data vector as it relates to trajectories of gentrification that feeds into the
legal realm as it relates to the private sector with community directed investment
which informs the resource management vector to implement investment decisions on
the land use system and different park types being established in certain areas (Figure
1). This study had surprising results in that, numbers and acreage of parks did not
predict gentrification. However, the research did a reveal strong lagged relationship
with the addition of park investment information. I found that gentrification was
associated with locations where the number and acreage of recreation and natural
passive parks increased from 1990 to 2000, and where high levels of park investment
were present. More research is needed to explore the causation of gentrification and
environmental gentrification. This research suggests that changes in access to and
investment in, differing park types (particularly natural passive and recreation parks)
may play a significant role in explaining social, economic, cultural, and in some
cases, racial changes to census block groups in Seattle. In the event that the Seattle’s
parks are significantly contributing to the processes of gentrification it is important
for community leaders to take action. The research was limited by the park
investment information used as it only pertained to community directed investment in
parks from one source of funding (the Department of Neighborhoods Matching fund).
Two steps should be taken in the future to create a more robust account of park
investments: (1) inclusion of all park investment including but not limited to – park
maintenance, nonmonetary investment (via volunteer clean up groups), and private
action groups, (2) analyzing a complete dataset of accepted and rejected grant
applications. This research deserves more attention regarding all park investment to
parks which includes both public funding as well as privatized investment. Including
all park investment may render more robust results about where money is being
150
invested and whether or not it is related to changes to economic and social
characteristics of neighborhoods. Moreover, analysis of privatized investment in
parks may speak to the social and cultural capital and organization that communities
may have over others. Secondly, analyzing a more complete database of public
grants/investment using both accepted and rejected grant applications would be useful
to identify if there was a disproportionate number of applications rejected in disparate
neighborhoods compared to applications being rejected in other neighborhoods (e.g
Wolch et al 2005). An analysis such as this one may reveal an institutionalized
systemic process of oppression that should be addressed in order ensure that equity is
being met for all communities.
This last chapter also creates an opportunity for further discussion and
research on managing park development/improvement decisions about park location,
access and investment to satisfy equity without promoting gentrification. As
discussed in Chapter 4, there is a call by Curran (2012) for communities to focus on
being “just green enough.” Environmental gentrification like traditional
gentrification can have the effect of displacing its residents and changing the social,
cultural, and economic dynamics of neighborhoods. Other research has shown that
when communities take on projects that are to large of a scale and do not consider the
immediate needs of the residents it can balloon and ultimately attract more attention
than it originally desired and the effect is displacement (Curran 2012). The study
presented on Seattle park investment as it relates to gentrification is the tip of the
iceberg and requires a great deal of research to thoroughly determine the effects of
community initiated investments in park development and improvement. However,
community leaders should consider implementing appropriately scaled projects that
are created by and for the residents of the community (Curran 2012). To enhance
future research on this topic, I suggest that an in depth case study be conducted on
various neighborhoods that completed community projects of different sizes and
assess whether or not it played a role in the area experiencing gentrification or not.
Overall, this research is lends itself to furthering the environmental justice
literature by reinforcing the importance of equitable distribution of environmental
goods. It also suggests that it may be important to distinguish between differing types
151
of parks to ensure that all community types have access. My research also informs
the gentrification literature by introducing a simplistic methodology for practitioners
to carry out in their respective cities. Further research and development activities
should be done to create a web application tool that can be used to quickly identify
whether or not a city or community is gentrifying or has gentrified or experiencing
perceptions of gentrification. Creating access to this type of information may help
individuals effectively plan and brace for impact within their community.